Skip to main content
NCBI home page
Neuro-Oncology logoLink to Neuro-Oncology
. 2011 Nov;13(Suppl 3):iii34–iii40. doi: 10.1093/neuonc/nor151

IMMUNOTHERAPY

PMCID: PMC3199174
Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-01. COMBINED CELLULAR IMMUNOTHERAPY AND SUICIDE GENE THERAPY MAXIMIZES ANTITUMOR RESPONSES IN EXPERIMENTAL MODELS OF METASTATIC BREAST CANCER IN THE CNS

Michelle J Hickey 1, Colin K Malone 1, Kate L Erickson 1, LE Gerschenson 2, Amy H Lin 1, Akihito Inagaki 1, Kei Hiraoka 1, Noriyuki Kasahara 1, Barbara Mueller 3, Carol A Kruse 1

Abstract

Cellular immunotherapy with alloreactive cytotoxic T lymphocytes (alloCTL) and suicide gene therapy using retroviral replicating vectors (RRV) have both been translated and are currently being evaluated clinically. Here we evaluated the feasibility and efficacy of combining these strategies using an experimental model of breast cancer metastatic to the brain. AlloCTL were made by one-way mixed lymphocyte-tumor reaction. Lymphocytes derived from healthy human donors were sensitized with MDA-MB-231 human breast cancer cells pre-exposed to IFNg to upregulate their HLA. The lymphocytes were HLA-mismatched to that on the tumor cells to ensure potent alloreactivity against the tumor cells but not normal brain neuroglia, which does not express class I HLA. AlloCTL preparations cultured for up to 14 days consisted primarily of CD3 + /CD8+ cells, which proliferated and produced proinflammatory IFNg, and displayed potent cytotoxicity when exposed to MDA-MB-231 target cells and the brain-tropic subline 231-BR in vitro and in vivo. Furthermore, alloCTL transduced with RRV coding for green fluorescent protein (RRV-GFP) or yeast cytosine deaminase (RRV-CD) were confirmed to efficiently confer RRV to the cancer cells in vitro and displayed motility on slides coated with 231-BR extracellular matrix protein extracts. Furthermore, RRV-CD transduced alloCTL and breast cancer cells were capable of converting the nontoxic prodrug 5-fluorocytosine (5-FC) into the toxic metabolite 5-fluorouracil (5-FU). Upon intracerebral implantation in immune-deficient mice, alloCTL and RRV-transduced alloCTL showed movement toward pre-established intracranial 231-BR xenograft tumors and induced their apoptosis. The individual and combined alloCTL and RRV-CD/5-FC therapies were efficacious in both subcutaneous and intracranial breast tumor xenograft models compared with various control groups. Peripheral tissues showed no significant systemic biodistribution of RRV, indicating the relative safety of the gene therapy approach. From these data, we conclude that strategies to combine cellular immunotherapy and gene therapy warrant further study for treatment of breast tumor foci metastatic to the brain.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-02. SUCCESSFUL SUPPRESSION OF A HUMAN GLIOMA XENOGRAFT USING IL13RA2-SPECIFIC MONOMERIC CHIMERIC ANTIGEN RECEPTOR

Seogkyoung Kong 1, Betty Tyler 2, Jinyuan Zhou 3, Bob S Carter 4, Henry Brem 2, Richard P Junghans 5, Prakash Sampath 1

Abstract

Glioblastoma multiforme (GBM) remains incurable because of tumor recurrence. The diffusely invasive tumor cell that disseminates throughout the brain is the hallmark of GBM. We developed and preclinically tested an IL13Ra2-specific monomeric chimeric antigen receptor (CAR) for clinical trials. We exploited IL13Ra2 as a GBM-specific tumor antigen because it is frequently overexpressed on a majority of GBM but not expressed on normal brain tissues. We hypothesized that monomeric CAR may be more suited for efficient killing of IL13Ra2 antigenic targets, which exist in monomeric form on GBM cells, because of the avoidance of steric hindrance that may occur when using a dimeric CAR. We developed monomeric CAR through mutations placed in the transmembrane domain and the hinge region to prevent dimerization. CD28 costimulation and monomeric CARs can be synergistic for productive IL13Ra2-specific T cell recognition through effective clustering of receptor-ligand pairs. We modified T cells using retroviral gene transfer to express a monomeric CAR comprising an extracellular IL13.E11K.R107K molecule linked to intracellular signaling components from the CD28 costimulatory molecule and the CD3 zeta molecule, redirecting T cells' killing specificity. We created IL13.E11K.R107K for selectively enhanced recognition and killing of GBM-associated IL13Ra2, discriminating from the physiologic shared IL13Ra1/ IL4R receptor. After in vitro expansion, genetically modified primary human IL13Ra2-specific T cells showed specifically enhanced recognition of IL13Ra2+ GBM cells, as demonstrated by efficient tumor cell killing, IL-2 & IFN-γ secretion, proliferation, and expansion. A single injection of IL13Ra2-specific T cells into the tumor-bearing rats resulted in a significant increase in survival in a human glioma xenograft model. These data provide evidence that direct, locoregional infusion of IL13Ra2-specific T cells into a glioma resection cavity can eradicate diffusely invasive GBM cells to provide an effective treatment for recurrent GBM.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-03. LONG-TERM FOLLOW-UP OF ACT III: A PHASE II TRIAL OF RINDOPEPIMUT (CDX-110) IN NEWLY DIAGNOSED GLIOBLASTOMA

Rose K Lai 1, Lawrence D Recht 2, David A Reardon 3, Nina Paleologos 4, Morris Groves 5, Myrna R Rosenfeld 6, Thomas Davis 7, Jennifer Green 7, Amy Heimberger 5, John Sampson 3

Abstract

BACKGROUND: Rindopepimut is a vaccine containing a 13–amino acid sequence unique to EGFRvIII, a constitutively activated mutation of epidermal growth factor receptor expressed in ∼31% of glioblastomas but absent in normal tissues, conjugated to KLH. ACT III was a multicenter, single-arm, phase II trial evaluating with temozolomide. As previously reported (Lai, SNO 2010), the primary endpoint was met, with 66% of the patients progression-free at 5.5 months from enrollment (∼8.5 months from diagnosis). Results were consistent with two previous phase II trials (ACTIVATE/ACT II). Long-term follow-up data from all phase II rindopepimut studies will be reported. METHODS: Sixty-five patients with newly diagnosed EGFRvIII–positive glioblastoma were enrolled after gross total resection and standard chemoradiation (∼3 months from diagnosis). The vaccine was administered biweekly three times, then monthly with temozolomide until progression. MGMT methylation status was assessed. RESULTS: Extended treatment with rindopepimut (up to ∼40 months) was well tolerated, with treatment-related adverse events consisting chiefly of immune-mediated reactions, such as local injection site and hypersensitivity reactions. Forty-five of 53 patients (82%) developed anti-EGFRvIII antibody titers >4 fold over baseline, and titers increased with duration of treatment. At a median follow up of 29.5 months, the median overall survival is 21.3 months (∼24.3 months from diagnosis) for all patients, 17.9 months (∼20.9 months from diagnosis) for unmethylated MGMT promoter, and 37.0 months (∼40 months from diagnosis) for methylated MGMT promoter. DISCUSSION: Despite previous reports suggesting reduced long-term survival for EGFRvIII+ patients (Pelloski 2007); >30% of patients treated with rindopepimut have survived >36 months from diagnosis. These median survival durations compare favorably to EGFRvIII+ patients matched for major entry criteria (15 months; Sampson 2010) and unselected EGFRvIII+ patients (12.6 months; Pelloski 2007). Tissue screening to further define outcomes for EGFRvIII+ patients treated with standard of care in other protocols is underway, while a definitive, international, randomized, placebo-controlled trial is planned to begin in 2011.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-04. WT1 IMMUNOTHERAPY FOR MALIGNANT GLIOMAS: RECURSIVE PARTITIONING ANALYSES AND PROGNOSTIC FACTORS

Naoya Hashimoto 1, Akihiro Tsuboi 1, Yasuyoshi Chiba 1, Noriyuki Kijima 1, Yoshihiro Oka 1, Manabu Kinoshita 1, Naoki Kagawa 1, Yasunori Fujimoto 1, Haruo Sugiyama 1, Toshiki Yoshimine 1

Abstract

We have conducted a phase II clinical trial of WT1 vaccination for recurrent gliomas (grades III and IV) and reported the result of glioblastoma with statistically minimal sample size, revealing the safety and effectiveness (Izumoto, Hashimoto, et al., J Neurosurg, 2008). Given the overall results of trial, we performed recursive partitioning analyses (RPA) (Carson, et al., J Clin Oncol, 2007) and looked for prognostic indicators. Patients were intradermally injected with an HLA-A*2402-restricted, modified 9-mer WT1 peptide every week for 12 weeks. Responses on MRI were analyzed by RECIST criteria 12 weeks after the initial vaccination. After the Kaplan-Meier analyses giving median progression-free survival (PFS) and overall survival (OS), 70 patients who were enrolled onto the trial were classified in RPA classes according to the method by Carson et al. In 51 glioblastoma patients, biological and radiological prognostic factors such as WT1 specific T cells in blood, delayed type hypersensitivity test (DTH), and responses on MRI-RECIST and methionine PET (T/N ratio) were compared statistically with survivals. Median OS in 7 classes (class I to VII) obtained by RPA showed the same tendency in terms of OS ranking, and 6 of 7 classes (class II to VII) presented better OS compared with Carson's. The only value lower than that of Carson's appeared in class I for unknown reasons. The hazard rate calculated for WT1 vaccination was 0.372, which was far better than those for other approaches in previously reported phase I and II studies. DTH-positive patients had significant prolongation of PFS, and the clinical response and DTH showed significant correlation (p = 0.0002). Methionine PET at 3 months after initial vaccination was also a strong prognostic indicator. RPA showed that WT1 vaccination is the most effective among the new drugs reported in Carson's paper. DTH and methionine PET can be important prognosticators of the response to WT1 immunotherapy.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-05. REGULATING IL-12:IL-23 BALANCE AFFECTS IMMUNOSUPPRESSIVE BEHAVIOR IN GLIOBLASTOMA

Suzanne M Birks 1, Michael Burnet 2, Geoff J Pilkington 1

Abstract

Glioblastoma are remarkably adept at evading immune responses mounted against them and have been shown to manipulate immune components into promoting tumorigenesis. They achieve this through a number of synergistic methods, including the release of inhibitory cytokines, that converge to inhibit T cell proliferation, activation, and migration and also induce T cell apoptosis. Administering compounds that upregulate the expression of tumor-suppressive cytokines while decreasing immunosuppressive cytokines may show therapeutic benefits. Using biopsy-derived cell cultures at P1-P2 as well as more established cell cultures at P10-P14, we have analyzed the macrophage and myeloid components of the population and assessed IL-10, IL-12, IL-23, and IFN-γ expression and secretion by Western blot, flow cytometry, and ICC. After treating the cells with 9 compounds (Synovo GmbH)—which regulate macrophage phenotype through NFκB activation and translocation promotion—at 10 µM for 72 hours, we performed the same analyses described above. Furthermore, we have assessed the influence of Stat3 activation and HIF1α on this phenomenon by carrying out Western blot analyses and siRNA knockdown. Using MAC387 antibody, we have shown the macrophage content of our biopsy-derived cells to vary between 30%–70% and similarly quantified the expression of IL-10, IL-12, IL-23, and IFN-γ. Studies show that several of the compounds enhance the expression of the tumor-suppressor cytokines IL-12 and IFN-γ while decreasing amounts of IL-10 and IL-23. Furthermore, we have shown that regulation of the IL-12:IL-23 balance may be related to Stat3 activation and that this trend appears to be most prevalent in the cell culture expressing high levels of the putative cancer stem cell marker CD133. Initial data suggest that modulating IL-12:IL-23 may promote the anti-tumor immune response. Further investigation is required of how this manipulation affects tumor cell behavior and survival and how the pathways described above may be better exploited for therapeutic potential.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-06. A CANCER VACCINE TARGETING CANCER STEM CELL ANTIGENS (ICT-107) DEMONSTRATES CORRELATED TUMOR ANTIGEN EXPRESSION AND PROGRESSION-FREE SURVIVAL AND MAY REDUCE THE CANCER STEM CELL POPULATION IN RECURRENT TUMORS

John S Yu 1, Christopher J Wheeler 1, Jeremy Rudnick 1, Mia Mazer 1, Hong Q Wang 1, Miriam A Nuno 1, Jaime E Richardson 1, Xuemo Fan 1, Jianfei Ji 1, Ray M Chu 1, James G Bender 2, Elma W Hawkins 2, Keith L Black 1, Surasak Phuphanich 1

Abstract

PURPOSE: This study evaluated the safety and immune responses to an autologous dendritic cell vaccine pulsed with class I peptides from tumor associated antigens (TAAs) expressed on gliomas and overexpressed in their cancer stem cell population (ICT-107). PATIENTS AND METHODS: TAA epitopes included HER2, TRP-2, gp100, MAGE-1, IL13Rα2, and AIM-2. HLA-A1– and/or HLA-A2–positive patients with brainstem glioma or glioblastoma multiforme (GBM) were eligible. Mononuclear cells from leukapheresis were differentiated into dendritic cells, pulsed with TAA peptides, and administered intradermally three times at 2-week intervals. RESULTS: Twenty-one patients were enrolled with 17 newly diagnosed GMS, three recurrent GBMs, and one brainstem glioma. Immune response data on 15 newly diagnosed patients showed that 47% were responders. TAA expression by qRT-PCR showed that all patient tumors expressed at least three TAAs, with 75% expressing all six. Correlations of increased PFS and quantitative expression of MAGE1, AIM-2, gp100, and HER2 were observed. A decrease in or absence of CD133 expression was seen in five patients who underwent reresection. In contrast, seven patients with recurrent GBM after radiation therapy and chemotherapy showed at least a 2.5-fold increase in CD133 expression in their recurrent tumors compared with their primary tumors. At last follow up, six of 16 newly diagnosed GBM patients showed no signs of tumor recurrence. Median progression-free survival in the newly diagnosed patients was 16.9 months. CONCLUSION: Expression of four ICT-107–targeted antigens in the prevaccine tumors correlated with prolonged progression-free and overall survival in newly diagnosed GBM patients. The goal of targeting tumor antigens highly expressed on GBM cancer stem cells is supported by the observation of decreased or absent CD133 expression in recurrent gadolinium-enhanced tumors. ICT-107 may thus immunologically target the cancer stem cell population of GBM. ICT-107 is presently undergoing testing in a multi-institutional randomized phase II trial.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-07. RESULTS OF A PILOT STUDY TO EVALUATE THE EFFECTS OF VACCINATIONS WITH HLA-A2-RESTRICTED GLIOMA ANTIGEN-PEPTIDES IN COMBINATION WITH POLY-ICLC FOR CHILDREN WITH NEWLY DIAGNOSED MALIGNANT BRAIN STEM GLIOMAS, NON-BRAINSTEM HIGH-GRADE GLIOMAS, OR RECURRENT UNRESECTABLE GLIOMAS

Ian F Pollack 1, Regina I Jakacki 1, Lisa H Butterfield 2, Hideho Okada 2

Abstract

Malignant astrocytomas of the brainstem and cerebral hemispheres have a poor prognosis despite current treatments, and new therapeutic approaches are needed. Based on our findings regarding antigen expression profiles of childhood gliomas, we initiated a pilot trial of subcutaneous vaccinations with synthetic peptides for glioma-associated antigen (GAA) epitopes emulsified in Montanide-ISA-51 every 3 weeks for 8 courses, and intramuscular administration of poly-ICLC in HLA-A2+ children with newly diagnosed brainstem glioma (BSG), cerebral high-grade glioma (HGG), or recurrent glioma. GAAs were EphA2, interleukin (IL)-13 receptor-α2, and survivin. Primary endpoints were safety and T cell responses against vaccine-targeted GAAs, assessed by ELISPOT and tetramer analysis. Treatment response was evaluated clinically and by MRI. To date, 18 children have been enrolled: 10 with newly diagnosed BSGs treated with irradiation, 4 with newly diagnosed HGGs treated with irradiation and concurrent chemotherapy, and 4 with recurrent gliomas. No dose-limiting toxicity has been encountered. One child with a BSG had transient tumor enlargement 4 months after beginning vaccination (7 months post-irradiation) that later regressed and culminated in a sustained partial response (PR), consistent with pseudoprogression. Two other children with BSG who had pseudoprogression and subsequent stabilization also remain alive > 1 year from diagnosis without further intervention. Principal toxicities have included injection site reactions and low-grade fevers, which have been mild. Among 16 patients evaluable for response, 12 had sustained stable disease, 1 had a PR, and 1 has a continuing complete response after surgery. Seven of 10 BSG patients have survived > 11 months after diagnosis. ELISPOT analysis, completed in 5 children to date, showed response to IL13Rα2 (n = 4), EphA2 (n = 2), and survivin (n = 1). Tetramer responses to IL13Rα2 and EphA2 were also noted. Our preliminary results demonstrate that a multipeptide vaccination approach in children with glioma is well tolerated and has evidence of immunological and clinical activity.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-08. SURGERY AND IMMUNOTHERAPY INCREASES SURVIVAL COMPARED WITH SURGERY ALONE FOR CANINE MENINGIOMAS

Matthew A Hunt 1, GE Pluhar 1, Brian M Andersen 1, Jose L Gallardo 1, Charles O Seiler 1, Karen S SantaCruz 1, John R Ohlfest 1

Abstract

Surgical resection is the mainstay of treatment of many meningiomas, and most are managed with local therapies. Systemic therapies have had limited effectiveness. We explored antitumor vaccination in dogs with spontaneous meningiomas. Historical mean survival in dogs treated with surgery alone is approximately 7 months. Thirteen dogs were treated with surgical resection followed by intradermal vaccination with autologous tumor lysate and immune adjuvant, either CpG oligodeoxynucleodtides (CpG ODN) (n = 5) or imiquimod (n = 8), every 2 weeks after surgery for 6 doses and evaluated by serial MRI. All 13 dogs have completed vaccination. Each dog had peripheral blood mononuclear cells (PBMC) and serum sampled serially. The sera were analyzed by Western blot against autologous tumor lysate. PBMCs were analyzed to determine the frequency of tumor-reactive T cells. Pathology demonstrated WHO grade III meningioma in 2 dogs and grade I meningioma in 11 with 2 of the grade I tumors showing evidence of brain invasion. Kaplan -Meier analysis of the immunotherapy/surgery groups (either CpG or imiquimod) compared with historical surgery alone controls (median survival 202 days) demonstrated significantly increased survival in the imiquimod group, with all dogs still alive with median follow-up of 247 days. No severe toxicity was seen. Western blot analysis demonstrated vaccine-induced antitumor antibody responses. PBMC analysis demonstrated CD8+ T cells with enhanced production of interferon gamma and increased cytotoxic degranulation compared with a prevaccination sample of PBMCs when analyzed by flow cytometry or ELISPOT in both the CpG ODN and imiquimod groups. Immunotherapy using autologous tumor lysate and immune adjuvant elicits tumor reactive IgG and T cell responses with either adjuvant and increases survival over historical controls when using imiquimod as the immune adjuvant. This approach could be applied to humans with meningiomas that have failed other standard therapies or those at high risk of recurrence.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-09. EFFICACY AND CHARACTERIZATION OF A GLIOMA VACCINE INCORPORATING AN IL-12–EXPRESSING SELECTIVELY REPLICATING ONCOLYTIC TYPE 1 HERPES SIMPLEX VIRUS

David F Bauer 1, Lawrence S Lamb 1, Daniel K Harmon 1, Xiaojia Zheng 1, Andrew K Romeo 1, George Y Gillespie 1, Jackie N Parker 1, James M Markert 1

Abstract

High-grade gliomas carry a poor prognosis despite advances in cytoreductive therapies. Immunotherapeutic approaches such as tumor vaccines may provide a powerful adjuvant to conventional therapies. We designed an irradiated whole-cell tumor vaccine to incorporate the oncolytic IL-12–expresing HSV-1 M002 in two mouse models, one incorporating the Neuro 2a neuroblastoma cell line and another incorporating the 4C8 glioma cell line. In the conventional model, Neuro2A and 4C8 tumors were placed into the right caudate nucleus of A/J and B6D2F1 mice, respectively, followed by intramuscular injection of vaccine or irradiated cell control at 7 and 14 days following tumor placement. There were no differences in survival between the vaccinated group and the control group that received Neuro 2A tumors; however, vaccinated mice with the slower-growing 4C8 tumors showed a significant survival advantage (p = 0.0028). A re-challenge of surviving mice with the same tumor placed in the contralateral hemisphere also showed a significant survival advantage over naïve control B6D2F1 tumor-bearing mice (p = 0.003). A prime/boost vaccination strategy was then incorporated in the Neuro 2A model, with mice receiving vaccine 7 days prior to tumor placement and another 7 days after tumor placement. Mice vaccinated using the prime-boost protocol showed a significant survival advantage (p = 0.0066) and were able to mount a successful immune response to a re-challenge of the same tumor placed in the contralateral hemisphere (p = 0388). An unrelated hepatocellular tumor cell line H6 grew unrestricted in vaccinated mice, suggesting that immunity to Neuro 2A tumors was specific. Sequential analysis of tumor-infiltrating lymphocytes revealed an early CD4+ T cell response, followed by an increasing CD8+ T cell response. In vitro cytotoxicity assays using splenocytes from vaccinated mice demonstrated specific killing of Neuro 2A tumor cells. These studies strongly suggest that a whole-cell tumor vaccine incorporating IL-12–expressing M002 HSV can produce a durable, specific immunization in murine models of intracranial tumors.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-10. PROPENTOFYLLINE DECREASES TUMOR GROWTH IN A RODENT MODEL OF GLIOBLASTOMA MULTIFORME BY A DIRECT MECHANISM ON MICROGLIA

Valerie L Jacobs 1, Russell P Landry 1, Joyce A De Leo 1

Abstract

Glioblastoma multiforme (GBM) is the most common and aggressive primary brain cancer, with a median survival of less than 1 year after diagnosis. The tumor microenvironment plays a critical role in tumor invasion and progression. Microglia and infiltrating macrophages are the most abundant immune cells in the tumor. In the present study, we demonstrate that systemic propentofylline (PPF), an atypical methylxanthine with CNS glial modulating and anti-inflammatory actions, significantly decreased tumor growth in a CNS-1 rat model of GBM by targeting microglia and not tumor cells. Rats received 1x105 CNS-1 tumor injections in the right striatum with daily intraperitoneal injections of PPF (50 mg/kg) or saline beginning the day of tumor injection. PPF did not cause apoptosis or decrease proliferation of CNS-1 tumor cells. Furthermore, we demonstrated in vitro that PPF decreased microglial migration toward CNS-1 tumor cells and decreased MMP-9 expression. The effects of PPF were shown to be specific to microglia and not peripheral macrophages. These results support a differential functional role of resident microglia and infiltrating macrophages in the brain tumor environment. Our data highlight microglia as a crucial target for future therapeutic development and present PPF as a possible drug for treatment of human GBM.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-11. ACUTE PAINFUL RADICULOPATHY AFTER INTRATHECAL RITUXIMAB

Jacoline E Bromberg 1, Jeanette Doorduijn 1, Joke W Baars 2, Guustaaf W van Imhoff 3, Roelien Enting 3, Martin J van den Bent 1

Abstract

Within a prospective phase II study (HOVON 80) of patients with recurrent diffuse large B-cell lymphoma in the CNS, three of 13 patients treated with intrathecal rituximab developed an acute, transient, extremely painful lumbosacral radiculopathy. All were treated with systemic R-DHAP every 4 weeks with intravenous HD-MTX on day 15 for three cycles. In addition, intrathecal rituximab was administered twice weekly via lumbar puncture. According to protocol, the first administration consisted of 10 mg of rituximab; thereafter the dose was increased to 25 mg. No patient experienced side effects from the first intrathecal administration of rituximab. However, after the first administration of 25 mg rituximab three of 13 treated patients reported extremely painful tingling sensations in the buttocks, legs, and feet immediately after administration and lasting 30-60 minutes. Concomitantly, a temporary increase in blood pressure was documented. Premedication with antihistaminics in the third patient was ineffective. No neurologic deficits occurred, and the pain resolved completely. The patients refused further treatment with intrathecal rituximab, and therapy was changed to intrathecal methotrexate without any side effects. After these events, the rituximab was diluted in saline to 5 mg/mL, the dose reduced to 10 mg per administration, and 4 mg intrathecal dexamethasone was administered concomitantly in all subsequent patients. Twelve additional patients were thus treated, and no further incidents of painful radiculopathy were observed. This serious, though completely transient, adverse effect of intrathecal rituximab precludes intrathecal administration of higher doses of rituximab via the lumbar route. This effect has never been described after intraventricular administration.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-12. AN IN VIVO IMMUNOTHERAPY SCREEN IDENTIFIES FC-MOX40L AS A POTENT DRUG CAPABLE OF CAUSING REGRESSION OF LARGE MURINE GLIOMAS

Katherine A Murphy 1, Jessica Bedi 1, Alan Epstein 2, John R Ohlfest 1

Abstract

Immunotherapy using vaccination shows promise as an adjuvant therapy for glioma patients. A significant hurdle in immunotherapy is the effect of anergy on immune cells, which results in functional inactivity and is often induced when antigen in recognized without the proper level of costimulatory signals. We hypothesized that addition of costimulatory molecules in combination with a tumor vaccine would result in a more robust immune response capable of eliminating an established tumor. C67BL/6 and B cell–deficient uMT mice bearing GL261 gliomas were vaccinated with GL261 tumor lysate mixed with an adjuvant, CpG oligodeoxynucleotides (ODN). The costimulatory ligands, OX40L, B7.1, CD137L, and GITRL, were fused to the Fc portion of the human immunoglobulin protein and administered in addition to the vaccine therapy. Lysate/CpG ODN vaccinations in combination with Fc-mOX40L yielded the best overall survival of the costimulatory ligands tested; significantly increasing survival of glioma bearing mice (p < 0.001). Seventy percent of the animals that received this treatment had complete tumor regression, and upon tumor rechallenge these animals were capable of clearing the tumor in less than a week. Treatment benefit was lost in the uMT mice. The combination of a mildly immunosuppressive regimen of temozolomide with vaccine plus OX40L cured 100% of mice bearing 7-day established tumors. OX40L monotherapy was able to extend the survival of (100%) or cure (20%) mice when treatment was started approximately 1 week before death on day 17. These data demonstrate OX40L has unique and incredibly potent activity against experimental gliomas relative to the other molecules tested. Cured mice developed immunological memory. In addition and contrary to other reports, this response is B cell–dependent, as all treatment benefit was lost in a B cell–deficient uMT model. The combined effect of temozolomide, vaccine, and Fc-mOX40L suggests that this could be a potent treatment option for patients.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-13. VACCINATION SITES ALTER THE T CELL PRIMING REQUIRED FOR A TUMORICIDAL RESPONSE AGAINST CNS TUMORS

Michael Olin 1, Brian Andersen 1, Lauryn Swier 1, John Ohlfest 1

Abstract

Cancer vaccines prime lymphocytes in the draining lymph nodes (DLNs) of the vaccination site. Many cancer vaccines are being developed for the treatment of glioma, yet there remains a lack of basic knowledge regarding the functional competence of DLNs in relation to the brain tumor site. We investigated how the injection site chosen for vaccination dictates CD8 T cell priming and animal survival in a GL261 glioma model. Glioma-bearing mice were injected subcutaneously in the neck or hind leg. All mice vaccinated in the neck died of tumor burden, while 37% of mice vaccinated in the leg had complete tumor regression. Interestingly, there was a disproportionate upregulation of CTLA4 on lymphocytes primed in the cervical lymph nodes (CLN) relative to the inguinal lymph node (ILN). To interrogate alterations in T cell priming, poly:ICLC plus ovalbumin (OVA) was administered in the neck, leg, or armpit for drainage into the axillary lymph node (ALN). Significantly more OVA-specific CD8 T cells were measured in blood in DLNs farther away from the brain in the following order: ILN > ALN > CLN. CD8 T cells primed in the leg had a more robust expression of CD44 and KLRG1 and lower CD127 relative to other sites, revealing a stronger effector phenotype following priming in the ILN. Moreover, there were significantly more OVA-specific CD8 T cells measured in the brain tumor site in mice vaccinated in the leg or armpit compared with those vaccinated in the neck. This effect appears specific to tumor burden because the injection site did not appreciably effect CD8 T cell priming in non–tumor-bearing mice. These data reveal a potentially critical aspect of study design, considering that past and previous clinical trials used a variety of injection sites. Our data imply that vaccination in the leg may increase priming of tumoricidal CD8 T cells capable of trafficking into the brain.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-14. TEMOZOLOMIDE-INDUCED LYMPHOPENIA ABOLISHED PRIMING OF CD8 T CELLS IN RESPONSE TO VACCINATION IN A MURINE MODEL

Adam J Litterman 1, David M Zellmer 1, John R Ohlfest 1

Abstract

The current standard of care in the treatment of glioblastoma multiforme requires chemoradiotherapy as well as adjuvant chemotherapy with the cytotoxic alkylating agent temozolomide. Both standard and dose-intensified regimens of temozolomide are associated with moderate to severe (grade II–grade III) lymphopenia, and it remains unknown what impact this lymphoablative chemotherapy has on concurrent administration of immunotherapies. Studies have shown that lymphopenia followed by adoptive transfer of T cells or bone marrow transplant increases transferred T cell expansion. A recent clinical trial reported that greater temozolomide-induced lymphopenia was associated with greater antibody and DTH responses to a peptide vaccine. We have used a vaccine combining a model antigen and adjuvant (Ovalbumin and PolyICLC) that rapidly induces a robust CD8 T cell expansion to examine the effect of temozolomide treatment on the generation of an endogenous immune response. Doses of temozolomide that mimic both the pharmacokinetic exposure and degree of lymphopenia observed in human patients significantly reduced the percentage and absolute number of antigen-specific CD8 T cells elicited by vaccination. In particular, the initial peak of expansion of effector CD8 T cells was reduced to essentially undetectable levels in mice with severe, temozolomide-induced lymphopenia. In mice bearing an ovalbumin-expressing glioma, temozolomide treatment, vaccination, and temozolomide followed by vaccination were all equally efficacious in reducing tumor burden. This lack of synergy between the chemotherapy and immunotherapy, combined with our observation of ∼20-fold lower levels of antigen-specific CD8 cytotoxic T lymphocytes in the blood of temozolomide pretreated mice, suggests that this temozolomide exposure does not enhance vaccination responses. While a lymphopenic milieu is associated with homeostatic proliferation, in practice this may be outweighed by the difficulty of priming endogenous T cell responses in lymphocytes that have been exposed to DNA-damaging drugs; future clinical trials examining antigen-specific T cell expansion will be needed to address this matter.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-15. PHASE 2 STUDY OF GENE-MEDIATED CYTOTOXIC IMMUNOTHERAPY ADJUVANT TO UP-FRONT SURGERY AND INTENSIVE TIMING RADIATION FOR MALIGNANT GLIOMA

EA Chiocca 1, Laura K Aguilar 2, Estuardo Aguilar-Cordova 2, Andrea G Manzanera 2, Kathryn R Harney 2, Jana Portnow 3, Benham Badie 3, Maciej Lesniak 4, Sue Bell 1, Abhik Ray-Chaudhuri 1, Balveen Kaur 1, Jayson Hardcastle 1, Robert Cavaliere 1, John McGregor 1, Simon Lo 1, Arnab Chakarvarti 1, John Grecula 1, Herbert Newton 1, Todd W Trask 5, David S Baskin 5, Pamela Z New 5

Abstract

BACKGROUND: Gene-mediated cytotoxic immunotherapy (GMCI) is an approach that uses AdV-tk, an adenoviral vector expressing the herpes virus thymidine-kinase gene, plus prodrug (GliAtak, Advantagene) to kill tumor cells and elicit an antitumor immune response. Preclinical studies demonstrated synergy with surgery and chemoradiation. A phase Ib study demonstrated the safety and potential efficacy of the combined approach. METHODS: A phase II study added 36 patients at the highest dose level of AdV-tk (3x1011 vp), delivered to tumor cells via tumor bed injection, followed by valacyclovir (Valtrex, GSK) for 14 days. Synergy with radiation therapy was maximized by starting radiation therapy 1 week after surgery to overlap with HSV-tk activity. RESULTS: Patient accrual was completed in November 2010. The median age was 57 years, KPS was 70-80 in 35% and 90-100 in 65%, and baseline MMSE was ≤ 26 in 50%, 27-29 in 25%, and 30 in 25%. Histology was glioblastoma multiforme (GBM) for 90% of patients, and 52% had subtotal resection. the median overall survival and 2-year overall survival rate are 15.8 months and 40% for all histologies and 15.8 months and 37.5% for GBM only. The median progression-free survival is 8 months. For patients with subtotal resection, the median overall survival is 14.3 months and 2-year overall survival is 34% compared with 20.7 months and 43% for patients with total resection. MGMT methylation status analysis is ongoing and currently available for 21 patients. Survival was not significantly different for the 12 patients with methylated MGMT (median 14.3 months, 2-year 41%) compared with the 9 patients with unmethylated MGMT (median 16.7 months, 2-year 28%). Health-related quality of life assessed with the FACT-Br patient-reported outcome measure was stable or showed improvement after treatment. CONCLUSIONS: Safety observed in the phase Ib trial was maintained with a larger patient population. The survival rate at 2 years may be increased compared with the standard of care alone, and results are particularly encouraging for patients with unmethylated MGMT.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-16. COMBINING ANTI–PD-1 (B7-H1) IMMUNOTHERAPY WITH STEREOTACTIC RADIOSURGERY IN A MOUSE ORTHOTOPIC GLIOBLASTOMA MODEL

Jing Zeng 1, Alfred P See 1, Jillian Phallen 1, Zineb Belcaid 1, Nicholas Durham 1, Christian Meyer 1, Emilia Albesiano 1, Gustavo Pradilla 1, Eric Ford 1, Hans Hammers 1, Betty Tyler 1, Henry Brem 1, Phuoc T Tran 1, Drew Pardoll 1, Charles G Drake 1, Michael Lim 1

Abstract

INTRODUCTION: Glioblastoma multiforme (GBM) cure rates remain low despite the best available therapies. Programmed-death-1 (PD-1) is a surface receptor that mediates T cell inhibition upon binding with its ligand PD-L1, expressed on many tumor types including human GBMs. Using a small-animal radiotherapy platform, we tested the combination of anti–PD-1 immunotherapy with stereotactic radiosurgery in a mouse orthotopic GBM model. METHODS: We performed intracranial implantation of mouse glioma cell line GL261 into C57/BL6 mice and stratified the mice into four treatment groups: control, radiation, anti–PD-1 antibody, and radiation plus anti–PD-1 antibody. Radiation was given on day 10 after implantation, under CT guidance using a 3-mm beam to 10 Gy. Anti–PD-1 antibody was given on days 10, 12, and 14 after implantation via intraperitoneal injection. RESULTS: Prolonged survival was demonstrated with combination anti–PD-1 therapy plus radiation compared with either modality alone: median survival durations were 25 days (control),. 27 days (anti–PD-1 antibody), 28 days (radiation), and 53 days (radiation plus antİ–PD-1 therapy) (p < 0.05, log-rank Mantle-Cox). Long-term survival was only seen in the combined treatment arm, with 20%–40% of animals alive at day 180+ after treatment. By day 21 after implantation, there was increased tumor infiltration by cytotoxic T cells (CD8 + /IFNγ + /TNFα+) and decreased regulatory T cells (CD4 + /FOXP3+) in the combined treatment group compared with the single modality arms. On day 90 after implantation, cured mice were rechallenged with flank tumors, with 0/8 visible tumors by 3 weeks, whereas naïve mice developed 8/8 tumors, each exceeding 1000 mm3 by 3 weeks. Repeating the survival experiment with CD8-depleted mice nullified the benefit from addition of anti–PD-1 therapy; whereas CD4-depleted mice continued to show a survival benefit with antibody therapy. CONCLUSIONS: The combination of PD-1 blockade and localized radiation results in long-term survival in mice with orthotopic brain tumors. These studies provide strong preclinical evidence to support combination trials in patients with GBM.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-17. GENERATION AND SCALE-UP OF POLYFUNCTIONAL CMV-SPECIFIC CYTOTOXIC T LYMPHOCYTES FOR ADOPTIVE IMMUNOTHERAPY OF GLIOBLASTOMA

Alexia Ghazi 1, Aidin Ashoori 1, Patrick Hanley 1, Vita Salsman 1, Donald R Schaffer 1, Zakaria Grada 1, Yvonne Kew 2, Suzanne Z Powell 2, Robert Grossman 2, Michael E Scheurer 1, Ann M Leen 1, Cliona M Rooney 1, Catherine M Bollard 1, Helen E Heslop 1, Stephen Gottschalk 1, Nabil Ahmed 1

Abstract

Glioblastoma multiforme (GBM) is the most common and the deadliest of primary brain cancers. Due to the failure of surgical and radiochemotherapy to improve GBM outcomes, novel biologically-based therapies are direly needed. We exploited the ubiquitous expression of CMV antigens pp65 and IE1-72 in GBM by generating polyfunctional autologous cytotoxic T lymphocytes (CTL) from GBM patients intended for adoptive transfer. We characterized CMV pp65 and IE1-72 antigen and CMV genome expression in sequentially diagnosed GBM patients. We generated and characterized the phenotype and functionality or CMV-specific CTL lines and scaled these up to clinically usable numbers. The phenotypes of the CTLs were determined using pentamer analysis for CMVpp65, and their functionality was tested using IFN-γ ELIspot and cytotoxicity assays. We demonstrate the successful generation and scale-up of polyfunctional CMV-specific CTLs that recognize and kill autologous GBM cells from GBM patients, and we demonstrate the feasibility of this methodology for adoptive transfer.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-18. DENDRITIC CELL VACCINE THERAPY FOR PATIENTS WITH RECURRENT GLIOBLASTOMA: A SINGLE-INSTITUTION POOLED ANALYSIS OF FOUR TRIALS

Jethro Hu 1, Chirag Patil 1, Miriam Nuno 1, Christopher Wheeler 1, Jeremy Rudnick 1, Surasak Phuphanich 1, Mia Mazer 1, Hong Qiang Wang 1, Ray Chu 1, Keith Black 1, John Yu 1

Abstract

INTRODUCTION: Clinical trials performed at our institution have shown that patients who receive dendritic cell immunotherapy are able to generate a tumor-specific response and that this response correlates with improved survival. However, each of these trials enrolled patients at varying stages of disease and tumors of varying grades. This pooled analysis was performed to evaluate the efficacy of dendritic cell immunotherapy for the subgroup of patients with recurrent glioblastoma. Prior analysis of patients with newly diagnosed glioblastoma demonstrated a median overall survival of 22.2 months and progression-free survival of 11.9 months. METHODS: From October 2000 to February 2011, 148 patients enrolled in 4 dendritic cell vaccine trials at our institution. Peripheral blood mononuclear cells were obtained by leukapheresis, differentiated into dendritic cells in culture, pulsed with either tumor peptides or autologous tumor lysate, and then administered intradermally over 3-4 doses. RESULTS: Over these 4 trials, 39 patients with recurrent glioblastoma were enrolled (32 at first recurrence, 4 at second recurrence, 3 at third recurrence), with median age 52 years (range 28-74 years) and median KPS 90 (range 60-100); 74.4% of patients were male. All patients underwent maximal surgical resection. Median survival from first vaccination was 10.9 months (range 9.6-17.4 months), with a 12-month survival rate of 43.2% (range 26.7%-58.8%) and a 24-month survival rate of 14.6% (range 4.8%-29.6%). Median progression-free survival was 2.8 months (range 1.9-4.0), with a 6-month progression-free survival rate of 17.1% (range 6.9%-30.9%). Treatment was well tolerated with no related grade 3/4 toxicities. As of this analysis, 10 patients are still alive and 3 patients are still progression-free. CONCLUSION: Efficacy of dendritic cell vaccine therapy for patients with recurrent glioblastoma is modest. Of note, 15 of the 39 patients were treated in the pre-bevacizumab era. Even so, median survival compares favorably to results obtained with bevacizumab at recurrence. We are currently participating in a multi-institutional phase II dendritic cell vaccine trial for patients with newly diagnosed glioblastoma.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-19. SYSTEMIC ANTITUMOR IMMUNITY TRIGGERED BY INTRATUMORAL IMMUNOMODULATION CURES DISTANT METASTATIC CNS LYMPHOMA

Aurelien Marabelle 1, Holbrook Kohrt 1, Joshua Brody 1, Richard Luong 1, Victor Tse 1, Ronald Levy 1

Abstract

BACKGROUND: We showed recently that intratumoral injections of CpG, a TLR-9 agonist, in association with systemic T cell immunomodulation using anti-CTLA4 and anti-OX40 monoclonal antibodies, can trigger an antitumor immune response able to cure distant subcutaneous lymphoma tumors in immunocompetent mice. SCIENTIFIC QUESTION: Based on this result, we asked whether this systemic antitumor immune response was able to cross the blood-brain barrier (BBB) and treat tumors in the CNS. RESULTS: We designed a CNS lymphoma model in immunocompetent mice and used bioluminescence imaging to assess the tumor growth in the brain without having to kill the mice. We have shown that injecting CpG in a subcutaneous lymphoma tumor, in combination with systemic anti-CTLA4/anti-Ox40 therapy, triggered an antitumor immune response that was able to cross the blood-brain barrier and eradicate disease of mice with established CNS lymphoma, including those with leptomeningeal metastases. Moreover, we have demonstrated that this antitumor immune response could be obtained by doing only local intratumoral immunomodulation using 100 fold less anti-CTLA4/anti-Ox40 antibodies than previously described. SIGNIFICANCE: Immunomodulatory antibodies are currently under clinical development for cancer treatment, and their major side effect so far is the triggering of autoimmune diseases. We have shown that injecting very little doses of these antibodies in combination with CpG at one tumor site is sufficient to trigger a systemic antitumor response able to eradicate distant sites including the CNS, which is usually considered a sanctuary site with immune privileges. IMPACT: We recently published the results of the phase I trial of intratumoral CpG in patients with follicular lymphoma. Anti-CTLA4 has just been FDA approved in melanoma patients, and anti-Ox40 antibodies are currently tested in phase I. Therefore, their combination shall be tested soon in patients with cancer.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-20. INTRACRANIAL THERAPY OF GLIOBLASTOMA WITH A NOVEL BISPECIFIC LIGAND-DIRECTED DIPHTHERIA TOXIN HUMAN EPIDERMAL GROWTH FACTOR AND UROKINASE-TYPE PLASMINOGEN ACTIVATOR RECEPTORS

Yan M Li 1, Huang Jun 1, Mousavi Shahryar 1, Vallera A Daniel 2, Hall A Walter 1

Abstract

INTRODUCTION: The prognosis for patients with glioblastoma multiforme (GBM) is poor despite maximal conventional treatments. We aimed to broaden the receptor recognition by immunotoxin to a wider range of GBM by incorporating two unrelated ligands, urokinase-type plasminogen activator (uPA) and epidermal growth factor (EGF), that recognize common GBM and neovascular antigens and to determine whether this immunotoxin could inhibit human GBM growth in our mouse intracranial model via convection enhanced delivery (CED). METHODS: DTAT, which contains the catalytic portion of diphtheria toxin for cell killing fused to the noninternalizing amino-terminal (AT) fragment of uPA, EGF was added to the same single chain protein. We tested this bispecific cytotoxin DTATEGF against human GBM cell lines. We established GBM intracranially in nude mice with U87-Luc or LN229-Luc cells genetically marked with a firefly luciferase gene. DTATEGF was delivered intracranially to treat mice with GBM via CED by a microosmotic pump. RESULTS: DTATEGF was highly potent and selective in killing GBM cells lines (U87, Ln229) and more potent than the monospecific cytotoxins. The growth inhibitory effects on glioblastoma cell lines by DTATEGF was accompanied by increased apoptotic population and expression of apoptotic proteins. In vivo, DTATEGF significantly reduced the intracranial tumor growth, shown by In Vivo Imaging System, compared with irrelevant immunotoxin control (P< 0.01). The mice with GBM treated with DTATEGF delivered by osmotic pump survived longer than the control (P < 0.01). No obvious local and systemic toxicity was seen in this mouse model of intracranial tumors. CONCLUSIONS: These studies show that a new cotargeting agent that simultaneously recognizes EGFR and uPAR is more effective for antiglioblastoma therapy because it simultaneously targets glioblastoma and neovasculature, and that DTATEGF delivered by CED has effective therapeutic advantages for glioblastoma and low toxicity in a mouse intracranial tumor model.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-21. PERIPHERAL NEUROPATHY ASSOCIATED WITH IPILIMUMAB: A REPORT OF TWO CASES

Iyavut Thaipisuttikul 1, Edward Avila 1

Abstract

We report two cases of severe peripheral neuropathy (PN) associated with ipilimumab, an anti–CTLA-4 monoclonal antibody. Both patients received a 10-mg/kg dose of ipilimumab every 3 weeks for treatment of melanoma. CASE 1: A 57-year-old man was treated with ipilimumab after a wide resection of melanoma. After the second dose, he developed rapidly progressive painful sensory polyneuropathy followed by weakness of distal lower extremities. Ipilimumab was discontinued. Magnetic resonance imaging (MRI) of the brain and spine were negative. Electrodiagnostic studies revealed severe sensorimotor polyneuropathy with demyelinating features. Multiple CSF studies showed persistent elevated WBC (22-94 cells/mm3) and elevated protein (68-95 mg/dL). Four CSF cytology studies were negative. Viral studies and paraneoplastic antibodies in serum and CSF were negative. His symptoms deteriorated rapidly despite trials of methylprednisolone, infliximab, and IVIg. Examination showed severe weakness with no movement below ankles, minimal strength in both hands and absent reflexes. He was treated with tacrolimus (0.3 mg/kg/day, adjusted per blood level) and methylprednisolone (1000 mg/day) for 10 days. Repeat CSF study showed normalization of WBC (8 cells/mm3) and protein (32 mg/dL) within 5 days of treatment. His symptoms stabilized and strength improved at a 2-month follow-up visit. CASE 2: A 62-year-old man developed lung metastasis 7 years after resection for melanoma, which was followed by interferon therapy. He did not respond to interleukin-2 and temozolomide and was treated with ipilimumab with an excellent response. After the fourth dose, he developed dysesthesias and decreased proprioception below the ankles but preserved strength and reflexes. He discontinued ipilimumab after the fifth dose due to acute hepatitis, which was treated with methylprednisolone and mycophenolate. His PN completely resolved within a few months. He was still disease-free 2 years after ipilimumab therapy. PN is an unusual adverse effect of ipilimumab. In these cases, coexistent CSF pleocytosis and hepatitis which improved with immunosuppression suggests an autoimmune phenomenon.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-22. DENDRITIC CELL VACCINES TARGETING CYTOMEGALOVIRUS IN GLIOBLASTOMA

Duane A Mitchell 1, Gary E Archer 1, Henry S Friedman 1, James E Herndon 1, Darell D Bigner 1, John H Sampson 1

Abstract

Glioblastoma multiforme (GBM) is the most common malignant primary brain tumor, and despite incapacitating conventional therapy, it remains universally fatal. The recent discovery and confirmation that GBMs, but not surrounding normal brain, express immunodominant CMV antigens provides an unparalleled opportunity to leverage these highly immunogenic viral proteins as tumor-specific immunotherapy targets. However, despite the attractiveness of targeting CMV in GBM, we and others have shown that patients with GBM are profoundly immunosuppressed, at least in part, as a result of an excessive number of immunosuppressive TRegs. TRegs are a phenotypically distinct population known to be uniquely dependent on the high-affinity IL-2 receptor (IL-2Rα, CD25) for their function and survival. Our previous investigations, however, demonstrate that the key features of this immunosuppressive phenotype can be completely reversed by eliminating TRegs. However, while monoclonal antibodies (MAbs) that block IL-2Rα can abrogate TReg function in animal models, they can also inhibit effective antitumor immune responses in mice and humans. Our preliminary data demonstrate that inhibition of IL-2Rα with blocking MAbs in normal mice undergoing TMZ-induced lymphopenia eliminates TRegs without impairing effector T cell immune responses to dramatically augment vaccine-induced immune responses. This differential effect of IL-2Rα blockade on TRegs and effector T cells also enhances the TEff:TRegs ratio, which has been shown to be an important predictor of antitumor efficacy and leads to enhanced efficacy in our established tumor treatment model. Of clinical relevance, a pilot study examining daclizumab-mediated anti–IL-2Rα blockade in patients with newly diagnosed GBM vaccinated with CMV pp65 mRNA transfected DCs during recovery from lymphodepletive TMZ reduces TReg frequency while permitting vaccine-stimulated antitumor effector cell expansion and progression-free survival interval of 27.4 months. Therefore, anti–IL-2Rα MAbs function differentially in nonlymphopenic versus lymphopenic contexts, and their use after treatment-induced lymphopenia may represent a novel and promising strategy for enhancing antitumor immunity.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-23. CYTOMEGALOVIRUS-SPECIFIC IMMUNE DEFICITS IN PATIENTS WITH GBM

Laura A Johnson 1, Gary E Archer 1, Smita K Nair 1, Robert Schmittling 1, Elizabeth Reap 1, John H Sampson 1, Duane A Mitchell 1

Abstract

Recent studies demonstrating that cytomegalovirus (CMV) antigens are expressed in a high proportion of glioblastoma multiformes (GBMs), without expression by adjacent normal brain, highlights these immunogenic viral antigens as potential immunotherapeutic targets. We examined the immunologic response to the immunodominant CMV antigen pp65 as well as other common acute and chronic viral antigens (influenza, Epstein-Barr Virus (EBV), adenovirus) in patients with newly diagnosed GBM compared with healthy age-matched normal donors to evaluate physiologic immunity against these viruses and to investigate the capacity to enhance CMV-specific immune responses using dendritic cell stimulation in vitro. Samples from patients with GBM exhibited normal frequencies of interferon (IFN)γ–positive CD4+ T cells after stimulation with CMV and other viral peptides in vitro. However, these same patients demonstrated a significantly reduced frequency and number of CMV-reactive CD8+ T cells, while maintaining normal responses to polyclonal stimulation with PMA and ionomycin and influenza and EBV antigens. In addition, antigen-specific responses were uniquely characterized in patients with GBM by high levels of IL-10 production, demonstrating an increased production of regulatory cytokines in responses to CMV antigenic stimulation. These data fit the characteristics of a phenomenon described as cellular exhaustion or anergy and suggest that patients with GBM may possess CMV-specific immune deficits in CD8+ T cells that allow for viral reactivation and persistence. Importantly, stimulation in vitro with autologous pp65 RNA-pulsed dendritic cells (DCs) led to the expansion of potent polyfunctional and cytolytic CMV-specific CD8+ T cells supporting the rationale for CMV-directed immunotherapy in patients with GBM. Ongoing studies in our laboratory aim to characterize the nature of CMV-immune deficits in patients with GBM and evaluate complementary strategies to DC stimulation to restore functional CMV reactive T cells in these patients.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-24. INTRACRANIAL ELIMINATION OF METASTATIC BREAST CANCER IN THE BRAIN USING LIGAND-TARGETED CYTOTOXIN AND CONVECTION ENHANCED DELIVERY IN A MOUSE XENOGRAFT MODEL

Yan M Li 1, Mousavi Shahryar 1, Huang Jun 1, Vallera A Daniel 2, Hall A Walter 1

Abstract

INTRODUCTION: Ten percent to 30% of patients with breast cancer have brain metastases; the median survival of these patients is 6-11 months. We aimed to use immunotoxins to target overexpressed receptors in breast cancer brain metastases and to determine whether these immunotoxins could inhibit the growth of metastatic breast cancer in the brain via convection enhanced delivery (CED). METHODS: Here we use a series of previously established or newly developed mono- and bi-specific immunotoxins containing the catalytic portion of diphtheria toxin (DT). DTAT utilizes the AT fragment of urokinase-type plasminogen activator (uPA) to target the uPA receptor (uPAR). DTAT13 targets both uPAR and the interleukin-13 receptor (IL-13R). DTEGF13 targets both IL-13R and the epidermal growth factor receptor (EGFR). DTATEGF targets both uPAR and EGFR. We performed assays including proliferation and cytotoxicity to test the immunotoxins' anticancer effect. We also established an intracranial model in nude mice with a breast cancer brain metastatic cell line genetically marked with a firefly luciferase reporter. Immunotoxins were delivered intracranially to treat the mice via CED by microsmotic pump. RESULTS: All the immunotoxins were highly potent in killing breast cancer brain metastatic MDA-MB-231-BR and CN-34-BR cells, accompanied by an increasing apoptotic population. In vivo, immunotoxins reduced intracranial tumor growth as shown by In Vivo Imaging System (P< 0.05). The mice with breast cancer brain metastases treated with immunotoxin delivered by osmotic pump survived longer than the control mice (P < 0.05), and several mice survived with a tumor-free status for 180 days. No obvious local and systemic toxicity was seen in our intracranial treatment experiments. CONCLUSIONS: These studies show that a group of mono- and bi-specific cytotoxins that recognize cancer specific receptors (EGFR, IL-13R, and uPAR) are effective in killing brain metastases from breast cancer. Immunotoxins delivered by CED have effective therapeutic advantages for treating breast cancer brain metastases and low toxicity in a mouse intracranial tumor model.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-25. RADIOSURGERY FOR MELANOMA BRAIN METASTASES IN THE IPILIMUMAB ERA: LONGER SURVIVAL IS POSSIBLE

Jonathan P Knisely 1, Harriet Kluger 2, Jaclyn Flanigan 2, Mario Sznol 2, James B Yu 2, Veronica L Chiang 2

Abstract

INTRODUCTION: A prospectively collected cohort of 77 patients who had definitive radiosurgery between 2002 and 2010 for melanoma brain metastases was retrospectively reviewed to assess the impact of ipilimumab use and other clinical variables on survival. METHODS: Seventy-seven patients with melanoma brain metastases treated with definitive radiosurgery at the Yale Gamma Knife Center were identified through an IRB-approved retrospective medical record review. Patient accrual spanned November 2002 to November 2010. Survival from the date of the first radiosurgical treatment was assessed in relation to age, gender, primary tumor location, whole brain radiation therapy use, performance status, and the use of ipilimumab. The diagnosis-specific graded prognostic assessment (DS-GPA) was calculated for each patient from the performance status and the number of brain metastases treated. RESULTS: The only significant differences between the 27 patients who received ipilimumab and those who did not were age at diagnosis of brain metastases and DS-GPA. Univariate and multivariate Cox proportional hazards analyses showed that DS-GPA score, independent of ipilimumab use, was the most significant predictor of overall survival (p < 0.001). A similar analysis with data censored at 24 months (performed to assess the short-term impact of ipilimumab use on survival) showed ipilimumab therapy was independently associated with an improvement in hazard for death (p = 0.03). CONCLUSIONS: The survival of patients with melanoma brain metastases managed with ipilimumab and definitive radiosurgery can exceed the expected 4-6 months. Using this treatment paradigm, the use of ipilimumab was associated with an increased median survival from 8.8 to 21.3 months with a 2-year survival rate of 19.7% vs. 47%, which remained significant even after adjustment for performance status.

Neuro Oncol. 2011 Nov;13(Suppl 3):iii34–iii40.

IM-26. DECITABINE IMMUNOSENSITIZES HUMAN GLIOMAS TO NY-ESO-1 T CELL RECEPTOR-TRANSDUCED LYMPHOCYTE TARGETING BY THE FAS/FAS LIGAND PATHWAY

Robert M Prins 1, Won Kim 1, Horacio Soto 1, Dominique N Lisiero 1, Dominique N Lisiero 1, Linda M Liau 1

Abstract

The lack of effective glioblastoma treatments remains a major health problem and highlights the need for novel and innovative approaches. Immunotherapy is an appealing strategy because of the potential ability for immune cells to traffic to and destroy infiltrating tumor cells. However, the absence of well-characterized, highly immunogenic tumor-rejection antigens (TRA) in glioblastoma has limited the implementation of targeted immune-based therapies. In this study, we tested whether the DNA methyltransferase inhibitor, decitabine, could influence expression of TRA on tumor cells, thereby facilitating surveillance by TRA-specific T cells. Treatment of human glioma cells with decitabine induced significant increases in the expression of NY-ESO-1 and other well characterized cancer testes antigens. This upregulation of NY-ESO-1 made these tumors susceptible to NY-ESO-1-specific T cell recognition and lysis. Interestingly, decitabine treatment of glioma cells also sensitized them to Fas-dependent apoptosis with an agonistic antibody, while a Fas-blocking antibody could largely prevent the enhanced functional recognition by NY-ESO-1–specific T cells. Such data support our hypothesis that agents that alter epigenetic cellular processes may immunosensitize tumor cells to tumor-specific T cell–mediated lysis. In summary, these results strongly suggest that decitabine treatment of human glioma induced the expression of TRA, rendering these cells susceptible to tumor-specific T cell recognition and killing. Thus, decitabine treatment can transform a previously nonimmunogenic glioma cell into a target that is efficiently recognized by NY-ESO-1–specific T cells.

Articles from Neuro-Oncology are provided here courtesy of Society for Neuro-Oncology and Oxford University Press

RESOURCES