Autologous glioma cell vaccine admixed with interleukin-4 gene transfected fibroblasts in the treatment of patients with malignant gliomas

Abstract

Background: The prognosis for malignant gliomas remains dismal. We addressed the safety, feasibility and preliminary clinical activity of the vaccinations using autologous glioma cells and interleukin (IL)-4 gene transfected fibroblasts.

Methods: In University of Pittsburgh Cancer Institute (UPCI) protocol 95-033, adult participants with recurrent glioblastoma multiforme (GBM) or anaplastic astrocytoma (AA) received gross total resection (GTR) of the recurrent tumors, followed by two vaccinations with autologous fibroblasts retrovirally transfected with TFG-IL4-Neo-TK vector admixed with irradiated autologous glioma cells. In UPCI 99-111, adult participants with newly diagnosed GBM or AA, following GTR and radiation therapy, received two intradermal vaccinations with the TFG-IL4-Neo-TK-transfected fibroblasts admixed with type-1 dendritic cells (DC) loaded with autologous tumor lysate. The participants were evaluated for occurrence of adverse events, immune response, and clinical response by radiological imaging.

Results and discussion: In UPCI 95-033, only 2 of 6 participants received the vaccinations. Four other participants were withdrawn from the trial because of tumor progression prior to production of the cellular vaccine. However, both participants who received two vaccinations demonstrated encouraging immunological and clinical responses. Biopsies from the local vaccine sites from one participant displayed IL-4 dose-dependent infiltration of CD4+ as well as CD8+ T cells. Interferon (IFN)-gamma Enzyme-Linked Immuno-SPOT (ELISPOT) assay in another human leukocyte antigen (HLA)-A2+ participant demonstrated systemic T-cell responses against an HLA-A2-restricted glioma-associated antigen (GAA) epitope EphA2883-891. Moreover, both participants demonstrated clinical and radiological improvement with no evidence of allergic encephalitis, although both participants eventually succumbed with the tumor recurrence. In 99-111, 5 of 6 enrolled participants received scheduled vaccinations with no incidence of major adverse events. Monocyte-derived DCs produced high levels of IL-12 p70. Treatment was well tolerated; however, we were unable to observe detectable IFN-gamma post-vaccine responses or prolonged progression-free survival in these participants.

Conclusion: Feasibility challenges inherent in the generation of a patient-specific gene transfection-based vaccine strongly suggests the need for more practical formulations that would allow for the timely administration of vaccines. Nevertheless, successful generation of type-1 DCs and preliminary safety in the current study provide a strong rationale for further efforts to develop novel glioma vaccines.

Figure 1

Type-1 polarized DCs are efficient producers of IL-12 p70. In UPCI 99-111, on each vaccination day, a small aliquot of DCs (8 × 10⁴) were tested for IL-12 production capability in comparison to DCs that were not stimulated with the type-1 cytokine-mixture (standard DCs). This was done with 24 hr stimulation of DCs (20 × 10³per well, duplicates) with CD40L-transfected J558 cells (50 × 10³per well). Supernatant was harvested and the production of IL-12 p70 was assayed by specific ELISA. Values indicate averages of duplicate samples. Bars indicate standard errors.

Figure 2

IL-4 gene transfected glioma cell vaccine elicited an IFN-γ response against EphA2 883–891 epitope. PBMC samples were obtained on days 1 (pre-vaccine on the day of the first vaccine), 8, 15 (on the day of the second vaccine), and 42, and saved as frozen cells until all these cells were thawed at the same time, cultured in the presence of 20 IU/ml hIL-2 and autologous glioma cells for 5 days, and evaluated for the frequency of IFN-γ-producing cells in response to T2 cells loaded with the HLA-A2-binding EphA2_883–891 peptide using ELISPOT assay. Each well contained 5 × 10⁴ CD8⁺ cells and each group was evaluated in triplicate. Specific IFN-γ spots were calculated by subtracting the average of control spots (triplicate variation within a group was less than 10% in non-peptide-loaded T2 cell groups) from the total numbers of spots in peptide-loaded groups. Values indicate averages of triplicate samples for each time point, and bars indicate standard deviations. The number of spots in each post-vaccine time point was at least three times the standard-deviation of the pre-vaccine value.

Figure 3

Regression of recurrent glioblastoma following vaccination with autologous glioma cells admixed with IL-4-transfected fibroblasts. Axial gadolinium-enhanced T1-weighted MR images obtained from the 2^nd participant in UPCI 95-033. Enhanced lesions were measured independently by a neuroradiologist.