Immunization with an antigen identified by cytokine tumor vaccine-assisted SEREX (CAS) suppressed growth of the rat 9L glioma in vivo.

We have reported previously that s.c. immunization of rats with IL-4 transduced 9L gliosarcoma cells (9L-IL-4) induced a potent antitumor immunity against intracranial, parental 9L tumors. Subcutaneous implantation of 9L-IL-4 influenced the systemic humoral response, which was demonstrated by Th2-type isotype-switching and the induction of cellular immune responses, which played a critical role in the rejection of tumors. Serological analyses of recombinant cDNA expression libraries (SEREX), has recently emerged as a powerful method for serological identification of tumor-associated antigens (TAAs) and/or tumor rejection antigens (TRAs). Because IL-4 is known to activate B cells and to promote humoral responses, and inasmuch as induction of humoral responses by central nervous system tumors has been reported to be minimal, we investigated whether the induction of a potent humoral immune response against 9L TAAs or TRAs in rats immunized s.c. with 9L-IL4 could be demonstrated. Screening of 5 x 10(5) independent clones of 9L-expression cDNA library for the presence of reactive antibodies in the serum from a 91-IL-4 immunized rat led to the identification of three different TAAs. One 9L TAA (clone 29) was demonstrated to be calcyclin, a member of the S-100 family of calcium-binding proteins. The second 9L TAA (clone 37) was demonstrated to be the rat homologue of the J6B7 mouse immunomodulatory molecule. The third TAA (clones 158 and 171) was determined to be the rat homologue of the mouse Id-associated protein 1 (MIDA1), a DNA-binding, protein-associated protein. Northern blotting demonstrated that message for calcyclin was overexpressed in 9L cells. Message encoding MIDA1 was highly expressed in parental 9L cells and thymus and, to a lesser degree, in testis, suggesting that MIDA1 was comparable with the cancer/testis category of TAAs. Sera obtained from animals bearing 9L-IL-4 were found to have a higher a frequency and titer of antibodies to these antigens when compared with sera obtained from rats bearing sham-transduced 9L (9L-neo) cells. To determine whether immunization with these TAAs induced antitumor immunity, animals were immunized by intradermal injection with expression plasmids encoding calcyclin or MIDA1. Subsequent challenge of rats with parental 9L resulted in significant suppression of tumor growth in animals immunized with MIDA1, but not with calcyclin. These results indicate that MIDA1 is an effective 9L TRA and will be useful for the investigation of specific antitumor immunity in this glioma model. Furthermore, these results suggest that this approach, termed "cytokine-assisted SEREX (CAS)," may serve as an effective strategy for identification of TRAs for in animal-glioma models of cytokine gene therapy, and potentially in humans undergoing cytokine gene therapy protocols as well.

Cytokine gene therapy of gliomas: induction of reactive CD4+ T cells by interleukin-4-transfected 9L gliosarcoma is essential for protective immunity.

Tumor cells genetically modified to secrete cytokines stimulate potent immune responses against peripheral and central nervous system tumors; however, variable results on the efficacy of this strategy for therapeutic intervention against established intracranial neoplasia have been reported. We have found that vaccination with rat 9L gliosarcoma cells expressing interleukin 4 (9LmIL4) induced a specific, protective, immune response against rechallenge with parental 9L tumors. In naive rats, sham-transfected 9L (9Lneo) tumors and 9LmIL4 tumors grew at comparable rates for 12-14 days, and then 9LmIL4 tumors regressed. After regression of 9LmIL4 tumors, rats were resistant to rechallenge with parental 9L cells. To investigate the mechanism(s) responsible for 9LmIL4-induced immunity, the phenotype and function of tumor-infiltrating lymphocytes (TILs) in 9Lneo and 9LmIL4 tumors were compared. In flow cytometric analyses, it was determined that CD4+ T cells were the predominant cell type in both 9Lneo and 9LmIL4 tumors at day 10. However, at the onset of regression (day 14), 9LmIL4 tumors were infiltrated predominantly by CD8+ T cells. To investigate functional aspects of the anti-9L tumor responses, we assessed the capacity of 9LmIL4 TILs to mediate specific lytic function or production of cytokines. In response to parental 9L, TILs isolated from day 14 9LmIL4 tumors were demonstrated to produce substantially greater amounts of IFN-gamma than did TILs from 9Lneo tumors. Although freshly isolated TILs from 9LmIL4 or control tumors did not lyse 9L cells in 51Cr-release cytotoxicity assays, specific cytotoxicity was demonstrable using TILs from day 14 9LmIL4 or splenocytes from 9LmIL4-bearing rats after their restimulation for 5 days with parental 9L tumor cells in vitro. Antibody blocking studies demonstrated that cytokine production and lytic activity by TILs, or splenocytes from 9LmIL4-immunized rats, were mediated in a T-cell receptor-dependent fashion. Because interleukin-4 also promotes humoral responses, quantity and isotype of immunoglobulins in sera from 9Lneo or 9LmIL4-immunized rats were compared. The amount of IgG1 antibodies was significantly increased in sera from 9LmIL4-immunized rats compared to sera from 9Lneo-bearing rats. Experiments using sublethally irradiated, naive rats adoptively transferred with splenocytes and/or sera from 9LmIL4-immunized or naive rats demonstrated that immune cells, with or without immune sera, protected recipients from challenge with parental 9L. Immune sera provided no protection when given with lymphocytes from naive rats, and it did not enhance protection against parental 9L when given in conjunction with lymphocytes for 9LmIL4-immunized rats. In additional adoptive transfer experiments, an essential role for CD4+ T cells in immunity was observed because their depletion from among splenocytes of 9LmIL4-immunized rats eliminated the protective effective against 9L, whereas depletion of CD8+ cells resulted in a more limited effect on protection against 9L. These data suggest that strategies for inducing systemic, long-term tumor-specific reactivity among CD4+ T cells will be critical for the development of immunotherapy of gliomas.

Characterization and transduction of a retroviral vector encoding human interleukin-4 and herpes simplex virus-thymidine kinase for glioma tumor vaccine therapy.

Vaccination with cytokine-transduced tumor cells represents a potentially important approach to the treatment of central nervous system tumors. We have recently demonstrated the therapeutic efficacy of tumor cell vaccines expressing the murine interleukin 4 (IL-4) and the herpes simplex virus-thymidine kinase in a rat brain tumor model in which nonirradiated vaccine cells can be eliminated by the subsequent administration of ganciclovir. In this report, we demonstrate the construction and characterization of a retroviral vector that encodes human IL-4, neomycin phosphotransferase, and herpes simplex virus-thymidine kinase genes for use in human clinical trials. An MFG-based retroviral vector was used to generate the recombinant retrovirus, TFG-hIL4-Neo-Tk, in which a long terminal repeat-driven polycistronic transcript encodes three cDNAs that are linked and coexpressed using two intervening internal ribosome entry site fragments from the encephalomyocarditis virus. The amphotropic retroviral vector TFG-hIL4-Neo-Tk was then used to infect human primary glioma cultures and skin-derived fibroblasts. After infection and G418 selection, cells produced 89-131 ng/10(6) cells/48 hours of human IL-4, which was determined to be biologically active. Transduced glioma cells were highly sensitive to the cytotoxic effect of ganciclovir. These data demonstrate the suitability of the TFG-hIL4-Neo-Tk vector for therapeutic studies of cytokine-transduced autologous tumor vaccination in patients with malignant gliomas.

Survival of patients with high grade glioma treated with intrathecal thiotriethylenephosphoramide for ependymal or leptomeningeal gliomatosis.

BACKGROUND: The diagnosis of leptomeningeal dissemination of malignant glioma (meningeal gliomatosis) is associated with poor survival. Intrathecal (IT) chemotherapeutic agents used to achieve tumor control and improve survival include methotrexate, cytosine arabinoside (ara-C), thiotriethylenephosphoramide (thio-TEPA), neocarzinostatin, and 3-[(4-amino-2-methyl-5-pyrimidinyl)methyl]-1-(2-chloroethyl)-1-nitros ourea hydrochloride (ACNU). Little information exists about survival following administration of IT chemotherapy. The authors report survival data from a series of patients with supratentorial anaplastic astrocytoma (AA) or glioblastoma multiforme (GBM) treated for ependymal or leptomeningeal gliomatosis with IT thio-TEPA. METHODS: The authors reviewed the records of 14 patients treated between 1991 and 1997 (GBM: n = 9; AA: n = 5). All patients were diagnosed with ependymal (n = 8) or leptomeningeal (n = 6) dissemination of tumor on the basis of clinical signs and symptoms, ependymal or leptomeningeal contrast enhancement on magnetic resonance imaging (MRI), and/or cerebrospinal fluid analysis. All 14 patients underwent placement of a ventricular reservoir system and subsequent instillation of IT thio-TEPA on a weekly basis for 6-12 weeks. Response to treatment was evaluated clinically and by MRI at intervals of 1-3 months and 3-6 months from the initiation of IT thio-TEPA. Data on survival from the time of diagnosis of dissemination was assessed. RESULTS: The median survival, from the time of diagnosis of ependymal or leptomeningeal dissemination, of patients who received IT thio-TEPA was 10 months (AA = 19 months; GBM = 10 months). Five of 14 patients had a radiographic response to treatment within 6 months. The median survival of patients with a radiographic response was 15.5 months, compared with 10 months for nonresponders. No significant neurotoxicity or myelopathy was observed. CONCLUSIONS: Early treatment with IT thio-TEPA may result in improved survival with minimal morbidity. Radiographic response may predict prolonged survival.

Brain tumor volume measurement: comparison of manual and semiautomated methods.

PURPOSE: To compare the reliability of two approaches to measuring enhancing brain tumor volumes--the conventional manual trace method and a threshold-based, semiautomated computer software method. MATERIALS AND METHODS: Two operators rated contrast material-enhanced, T1-weighted axial magnetic resonance (MR) image data sets from 16 patients aged 21-71 years with high-grade gliomas. Each MR data set was rated twice by using manual tracing and twice by using the semiautomated method. The semiautomated measurement method involved a thresholding algorithm based on mixture modeling. The data collection time for each method was recorded. Reliability was measured by using inter- and intraoperator agreement indexes. RESULTS: Mean intraoperator agreement indexes (+/- SD) were 0.90 +/- 0.09 (operator 1) and 0.83 +/- 0.15 (operator 2) for the manual trace method and 0.83 +/- 0.17 (operator 1) and 0.84 +/- 0.16 (operator 2) for the semiautomated measurement method. The mean interoperator agreement was 0.85 +/- 0.14 for the manual method and 0.82 +/- 0.18 for the semiautomated method. The semiautomated method was faster than the manual trace method by an average of 4.6 minutes per patient. CONCLUSION: The semiautomated computer method of measuring tumor volume was faster than the manual trace method. Semiautomated computer approaches offer an alternative to manual tracing for measuring serial tumor volumes in patients with high-grade brain neoplasms.

Effective cytokine gene therapy against an intracranial glioma using a retrovirally transduced IL-4 plus HSVtk tumor vaccine.

To explore the potential for molecular immunotherapies in the treatment of malignant gliomas, we evaluated the efficacy of subcutaneous tumor cell vaccines in the treatment of intracranial 9L tumors, using 9L gliosarcoma cell lines stably transduced with the murine interleukin-4 cDNA (9L-IL4), the herpes simplex virus-thymidine kinase cDNA (9L-Tk) or both (9L-IL4-Tk). The expression of multiple genes from a single transcript was achieved by incorporating internal ribosomal entry site (IRES) cassettes in the retroviral constructs. Subcutaneous immunization of rats with nonirradiated 9L-IL4 cells or 9L-IL4-Tk cells followed by treatment with ganciclovir (GCV) completely protected the animals from a subsequent intracranial challenge with wild-type 9L cells. In contrast, only 50% of animals immunized with 9L-Tk cells and 0% of 9L-neo immunized animals rejected the same challenge with wild-type 9L. More importantly, treatment of established (day 3) intracranial 9L tumors with genetically engineered tumor cells resulted in long-term survival (> 100 days) for 25-43% of 9L-IL4-Tk immunized animals and for 27% of nonirradiated 9L-IL4 immunized animals. In striking contrast, no 9L-Tk, 9L-neo or irradiated 9L-IL4 immunized animals survived for more than 33 days. As a marker of a cellular immune response, splenocytes from nonirradiated 9L-IL4, 9L-Tk or 9L-IL4-Tk immunized animals produced interferon-gamma (IFN-gamma) in greater amounts than those from 9L-neo immunized or Hank's balanced salts solution (HBSS) treated animals when stimulated with wild-type 9L in vitro. Our findings support the use of tumor cell vaccines expressing the IL-4 and HSVtk genes for the treatment of malignant gliomas.

Phase II study of phenylacetate in patients with recurrent malignant glioma: a North American Brain Tumor Consortium report.

PURPOSE: To determine the response rate, time to treatment failure, and toxicity of phenylacetate in patients with recurrent malignant glioma and to identify plasma concentrations achieved during repeated continuous infusion of this agent. PATIENTS AND METHODS: Adult patients with recurrent malignant glioma were treated with phenylacetate. The schedule consisted of a 2-week continuous, intravenous infusion followed by a 2-week rest period (14 days on, 14 days off). A starting dose of 400 mg/kg total body weight per day of phenylacetate was initially used and subsequently changed to 400 mg/kg/d based on ideal body weight. Intrapatient dose escalations were allowed to a maximum of 450 mg/kg ideal body weight/d. Tumor response was assessed every 8 weeks. The National Cancer Institute common toxicity criteria were used to assess toxicity. Plasma concentrations achieved during the patients' first two 14-day infusions were assessed. RESULTS: Forty-three patients were enrolled between December 1994 and December 1996. Of these, 40 patients were assessable for toxicity and response to therapy. Reversible symptoms of fatigue and somnolence were the primary toxicities, with only mild hematologic toxicity. Thirty (75%) of the 40 patients failed treatment within 2 months, seven (17.5%) had stable disease, and three (7.5%) had a response defined as more than 50% reduction in the tumor. Median time to treatment failure was 2 months. Thirty-five patients have died, with a median survival of 8 months. Pharmacokinetic data for this dose schedule showed no difference in the mean plasma concentrations of phenylacetate between weeks 1 and 2 or between weeks 5 and 6. CONCLUSION: Phenylacetate has little activity at this dose schedule in patients with recurrent malignant glioma. Further studies with this drug would necessitate an evaluation of a different dose schedule.

Bone marrow-derived dendritic cells pulsed with a tumor-specific peptide elicit effective anti-tumor immunity against intracranial neoplasms.

Although the central nervous system (CNS) is often regarded as an immunologically privileged site, it is well established that specific CNS immunoreactivity can be generated through peripheral vaccination with CNS antigens. Dendritic cells (DC) are potent antigen presenting cells of hematopoietic origin that have emerged as a promising tool for cancer immunotherapy capable of evoking significant anti-tumor immunity when pulsed with tumor-associated peptides. To explore a role for DC-based immunization strategies for the treatment of CNS tumors, we developed a brain tumor model using the C3 sarcoma cell line which expresses the tumor-specific, major histocompatibility complex (MHC) class I-restricted peptide epitope E7(49-57). Syngeneic C57Bl/6 mice receiving intravenous (i.v.) injections of bone marrow-derived DCs pulsed with E7 peptide were effectively protected against a subsequent intracerebral challenge with C3 tumor cells. More importantly, this systemic immunization strategy was effective in a therapy model as 67% of animals (10 of 15) with established (day 7) intracerebral C3 tumors treated with 3 weekly injections of E7 peptide-pulsed DCs achieved a long-term survival (>90 days) while no control animals survived beyond day 41. In vivo depletion of CD8+ cells, but not CD4+ or asialo-GM1+ cells, abrogated the efficacy of E7 peptide-pulsed DC therapy of established tumors, indicating a pivotal role of specific CD8+ T-cell responses in mediating the anti-tumor effect. Our findings support the hypothesis that effective CNS anti-tumor immunoreactivity can be generated with DC-based tumor vaccines.

Phase I study of paclitaxel in patients with recurrent malignant glioma: a North American Brain Tumor Consortium report.

PURPOSE: To determine the maximum-tolerated dose (MTD) of paclitaxel administered as a 3-hour infusion in patients with recurrent malignant glioma. PATIENTS AND METHODS: Patients were stratified by starting dose of paclitaxel and concurrent anticonvulsant (AC) use and were treated in cohorts of three patients. The starting dose was 240 mg/m2 administered intravenously with escalations of 30 mg/m2 until the MTD was established. Pharmacokinetic data were obtained for each patient for the first infusion. Tumor response was assessed at 6-week intervals and treatment was continued until documented tumor progression, unacceptable toxicity, or a total of 12 paclitaxel infusions. RESULTS: From April 1995 to December 1996, 34 patients were treated; 27 patients in the AC group and seven patients in the non-AC group. The MTD for patients who received ACs was established at 360 mg/m2 and the dose-limiting toxicity (DLT) was central neurotoxicity, characterized as transient encephalopathy and seizures. In contrast, the MTD for patients who did not receive ACs was 240 mg/m2, and myelosuppression, gastrointestinal toxicity, and fatigue were the DLTs. Pharmacokinetic data confirmed that the plasma drug levels and clearance rates were similar for patients in both groups at the respective dose levels that produced DLTs. CONCLUSION: The pharmacokinetics of paclitaxel are altered by ACs, and significantly larger doses of the drug can be administered to patients with brain tumors on AC therapy. The toxicity profile is different for patients on AC therapy treated at these higher doses. A phase II study has been initiated that uses a dose of 330 mg/m2 for patients on AC therapy and 210 mg/m2 for patients not on AC therapy.

Proliferation index as a predictor of prognosis in malignant gliomas of childhood.

BACKGROUND: The prognosis for children with high grade gliomas remains somewhat unpredictable because histologic features alone provide an imperfect assessment of the biologic behavior of a given lesion. Whereas some patients experience prolonged disease control after surgery and adjuvant therapy, others with lesions that appear comparable exhibit rapid disease progression and death. METHODS: Because proliferative activity may provide a potential correlate of biologic aggressiveness, the authors examined the relationship between MIB-1 labeling index and outcome in a series of 29 archival pediatric malignant nonbrainstem gliomas from patients treated consecutively at the study institution between 1975 and 1992, in which clinical, histologic, diagnostic, and therapeutic parameters were previously defined. Three patients who died perioperatively were excluded from outcome analyses. All tumors were rereviewed by two neuropathologists and classified as Grade 3 or 4 lesions based on contemporary guidelines. RESULTS: Among the specimens from the 26 patients who survived the perioperative period, a striking difference in outcome was apparent between tumors with MIB-1 indices < 12 (n = 10) and those with indices > 12 (n = 16). Median progression free survival was >48 months in the low MIB-1 group compared with only 6 months in the high MIB-1 group (P = 0.014, rank-sum test). Median overall survival was >48 months in the low MIB-1 group compared with only 16 months in the high MIB-1 group (P = 0.012). MIB-1 index remained associated with survival after taking into account the effect of resection extent, which also correlated strongly with outcome in this cohort. Although MIB-1 index was associated with histopathologic grade (Grade 3: 11.9 +/- 9.7 vs. Grade 4: 27.3 +/- 19.0; P = 0.015, Fisher's exact test), it proved to be a much stronger predictor of outcome than histology. CONCLUSIONS: MIB-1 index may supplement routine histologic classification as a means for improving the accuracy of predicting the biologic behavior of childhood malignant gliomas and may provide a basis for stratifying patients in future malignant glioma studies and refining therapeutic decision-making.

The relationship between TP53 mutations and overexpression of p53 and prognosis in malignant gliomas of childhood.

The prognosis for children with high-grade gliomas remains somewhat unpredictable. Although prolonged disease control is sometimes achieved after surgery, radiotherapy, and chemotherapy, most patients exhibit rapid disease progression. Because p53-dependent apoptosis mechanisms are involved in the cytotoxic effects of irradiation and chemotherapy, we questioned whether p53 status might be associated with outcome in childhood malignant gliomas. Therefore, we examined p53 status, both immunohistochemically and by direct sequencing of exons 5-8, in a series of 29 archival pediatric malignant non-brainstem gliomas treated consecutively at our institution between 1975 and 1992. Eighteen tumors had dense p53 staining in the majority of cells, although only 11 had mutations of the p53 gene (TP53). On univariate analysis, there was a significant association between p53 overexpression and a shorter progression-free survival (PFS) and overall survival (OS; P = 0.019 and 0.013, respectively; rank sum test). In addition, there was a significant association between TP53 mutations and a poorer PFS (P = 0.04), and a strong trend toward a shorter OS among patients with TP53 mutations (P = 0.06). Median PFS and OS for patients with TP53-mutated tumors were 6 months and 16 months, respectively, and for those with p53 overexpression 5.5 months and 14 months, respectively, versus 16 months and 25 months, respectively, for those without TP53 mutations and 25 months and >4 years, respectively, for those without p53 overexpression. The percentage of patients in this series with TP53 mutations (37.9%) was substantially higher than in previous studies of childhood gliomas and comparable to the frequency of mutations noted in adult gliomas. However, both TP53 mutation and p53 overexpression were significantly less frequent in tumors from children younger than 4 than from older children (P = 0.02 and 0.01, respectively). These results indicate that p53 mutation and expression status may be associated with prognosis in childhood malignant gliomas, and thus may provide a basis for stratifying patients biologically in future malignant glioma studies.

NKR-P1+ cells localize selectively in Rat 9L gliosarcomas but have reduced cytolytic function.

To better understand immune responses to brain tumors and to develop possible approaches for immunotherapy, we have investigated the leukocyte populations infiltrating the rat 9L gliosarcoma. By immunocytochem-ical analyses of the cells infiltrating the tumor, we observed a substantial number of cells expressing natural killer cell receptor protein 1 (NKR-P1), a marker expressed only on rat lymphocytes capable of non-MHC-restricted cytotoxicity. Previous investigations have determined the existence of three populations of NKR-P1+ lymphocytes in normal rats, including NKR-P1bright/T-cell receptor (TCR)-/CD3-/CD5- (approximately 5-15%), NKR-P1dim-/TCRalphabeta+/CD3+/CD5+ (approximately 1-5%), and NKR-P1dim/TCRgammadelta+/CD3+/CD5+ (approximately 0.5-2%). By one-parameter flow cytometry, it was determined that NKR-P1+ cells constituted 30-60% of the lymphocytes in 9L tumors. Among splenic lymphocytes or peripheral blood leukocytes, NKR-P1bright cells are 1.5-4.5 times more numerous than NKR-P1dim cells. In striking contrast, NKR-p1dim cells were 4-5 times more numerous than NKR-P1bright cells among lymphocytes isolated from 9L tumors. Using quantitative analyses of laser confocal microscopic scans, we determined that NKR-P1dim cells were approximately 4 times as numerous as NKR-P1bright cells in situ, confirming flow cytometric findings. By two-color now cytometric analyses, it was observed that approximately 5-10% of the cells were NKR-p1bright/CD5-/TCR-, a phenotype representative of NK cells. Also, approximately 11-25% of the cells were NKR-P1dim/CD5+/TCR+ cells, corresponding to the T-cell subset with non-MHC-restricted lytic function. In addition, we observed a cell population among 9L-derived lymphocytes with a NKR- p1dim/CD5-/TCR- phenotype (approximately 15-25%). Cells of this phenotype have not been reported previously, and most likely represent NK cells down-modulated for expression of NKR-P1. Alternatively, they might represent cells of unknown origin or cells down-modulated for expression of T-cell markers in the microenvironment of 9L tumors. We also compared the lytic capacity of NKR-P1+ populations derived from normal animals and from 9L gliosarcomas. In these experiments, it was determined that, although cells isolated from 9L tumors had some capacity to lyse tumor target cells, they were clearly less efficient than cells isolated from normal splenocytes. Cumulatively, these data suggest that there is selective localization of cells capable of mediating antitumor responses in 9L, but that tumor-associated factors may down-regulate their function and expression of NKR-P1.