α-type-1 polarized dendritic cell-based vaccination in recurrent high-grade glioma: a phase I clinical trial

Nine patients with recurrent high- grade glioma were enrolled in a phase I clinical trial of a peptide cocktail-pulsed α-type1-polarized DC-based vaccine approved by the Institutional Review Board of the Shizuoka Cancer Center, Shizuoka, Japan. All patients gave written informed consent. Patients` characteristics are listed in Table 1. Inclusion criteria were: i) histological diagnosis of high-grade glioma (WHO grade III or IV) [19], ii) poor response to standard treatment [surgical therapy (ST) + radiation therapy (RT) + temozolomide], iii) 20 y.o. ≦ age < 80 y.o., iv) performance status ≦ 2, v) evaluable lesions in imaging diagnosis, vi) HLA genotyping ; HLA-A2 or A24, vii) no severe systemic organ dysfunction, viii) no severe hematostatic dysfunction, ix) informed consent obtained from patients, and x) life expectancy > 6 months. Minimum doses of corticosteroid (dexamethasone up to 1 mg/day) were permitted for patients with neurological deficits due to mass effects by the lesions.

Exclusion criteria were : i) severe systemic infection and organ failure, ii) pregnancy, iii) severe immunological disorders (autoimmune disease, immunosuppression), iv) multiple cancers, and v) anaphylaxis to synthetic peptides. All the patients received intradermally (i.d.) 4 DC vaccines at the posterior neck weekly, and toxicity was checked. DCs were injected in a dose-escalation design at a dose level per cohort of 1.0, 2.0, and 5.0 ×10⁷/body/shot. The injected DC number was calculated from the percentage of Lin^-HLA-DR⁺ gated populations in a FACS analysis.

The procedures for preparing the DC vaccine were performed within a high efficiency particle arrestance (HEPA)-filter clean-air barriered good manufacturing practice (GMP) cell processing facility. A standard operation procedure (SOP) for DC vaccine production was established according to institutional GMP-based guidelines. The procedures have been described before [20]. Briefly, leukapheresis products were washed and centrifuged using density-adjusted OptiPrep™ (Axis-Shield PoC, Oslo, Norway), and the monocyte layer at the top was retrieved. On day1, cells were transferred to an X-fold culture bag (Nexell, Irvine, CA) and cultured in the presence of GM-CSF at 50 ng/ml (CellGenix, Freiburg, Germany) and IL-4 at 50 ng/ml (CellGenix) in X-VIVO15 serum-free medium (Biowhittaker, Walkersville, MD). On day6, cells were activated by the addition of TNF-α at 10 ng/ml (CellGenix), IL-1β at 10 ng/ml (CellGenix), IFN-α at 3000 U/ml (Dainippon Sumitomo Pharma Co. Ltd, Osaka, Japan), IFN-γ at 1000 U/ml (Shionogi & Co. Ltd, Osaka, Japan), and poly I/C at 20 μg/ml (Amersham Biosciences Corp., Piscataway, NJ). On day8, harvested cells were pulsed with a cocktail of 5 synthetic peptides (25 μg/ml each) restricted to HLA A2 or A24 and KLH (25 μg/ml, Intracell, Frederick, MD). Finally, DC-enriched cells were washed and cryopreserved in Cryocyte bags (Baxter Healthcare Co., Deerfield, IL) until used. The purity of CD14⁺ cells was evaluated with a flow cytometer (FACSCalibur, Becton-Dickinson Co., CA) before and after OptiPrep™ separation. The percentage of DCs was rated as the lin^-HLA-DR⁺ population (lineage antibodies including CD3, CD14, CD16, CD19, CD20, CD56 ; BD Biosciences, San Jose, CA). The frequencies of the DC-related markers were determined using various antibodies for CD1a, CD11c, CD33, CD40, CD54, CD80, CD86, CD123, CD205, CD207, CMRF44, CMRF56, E-cadherin CCR7, HLA-class I, and HLA-DR (BD Biosciences). The ratio of DC1 (CD11c⁺HLA-DR⁺) and DC2 (CD123⁺HLA-DR⁺) was calculated using a flow cytometric analysis reported by Ferrari et al. [21].

The following peptides restricted to HLA-A2 or A24 were synthesized according to GMP standards by Multiple Peptide Systems, CA: HLA-A2: WT1126-134 (RMFPNAPYL), WT1235-243 (CMTWNQMNL), gp100209-217 (IMDQVPFSV), HER2369-377 (KIFGSLAFL), MAGE-A3271-279 (FLWGPRALV) ; HLA-A24: WT1235-243M (CYTWNQMNL), WT1235-243 (CMTWNQMNL), HER263-71 (TYLPTNASL), MAGE-A1135-143 (NYKHCFPEI), MAGE-A3195-203 (IMPKAGLLI).

Cultured mature DCs were collected, and plated in a round-bottomed 96-well microplate at 5 × 10⁴ cells/well. To stimulate IL-12p70 production by DCs, a CD40 ligand-expressing mouse plasmacytoma cell line, J558, was added at 1 × 10⁵/well. CD40L-transfected J558 cells were kindly supplied by Dr. Kalinsky, the University of Pittsburgh Cancer Institute, Pittsburgh, PA. Both cells were incubated for 24 hrs. Finally, supernatants were collected and IL-12p70 levels were measured using an ELISA kit specific for human IL-12p70 (Endogen, Woburn, MA).

High-grade glioma tissues were obtained from patients who gave written informed consent. The expression of tumor antigens including MAGE-A1, A3, WT-1, HER2 and gp100 was investigated using a non-quantitative RT-PCR and an immunohistochemical (IHC) analysis as described previously [22]. The primer sequences used were 5’-CGGCCGAAGGAACCTGACCCAG-3’ and 5’-GCTGGAACCCTCACTGGGTTGCC-3’ for MAGEA-1, 5’-TGGAGGACCAGAGGCCCCC-3’ and 5’-GGACGATTATCAGGAGGCCTGC-3’ for MAGE-A3, 5’-TCTTCAGAGGCATTCAGGATGTGC-3’ and 5’-GAGTCCTGGTGTGGGTCTTCAGGT-3’ for WT-1, 5’-CTATGGCTGCCTCTTAGACCATGTCC-3’ and 5’-AAAGTCATCAGCTCCCACACAGTCAC-3’ for HER2, 5’-TAGTGGAACCCTGATCTCTC-3’ and 5’-TCAGGGATAGGTAGCTCTCT-3’ for gp100, and 5’-GGCTACAGCTTCACCACCAC-3’ and 5’-GTACTTGCGCTCAGGAGGAG-3’ for beta-actin. HLA-class I protein expression was also evaluated using an IHC analysis. Antibodies against MAGE-A1 (Thermo scientific, Flemont, CA), MAGE-A3 (Abnova, Taipei, Taiwan), WT-1 (DakoCytomation, Glostrup, Denmark), HER1 (DakoCytomation), gp100 ((DakoCytomation) and human HLA class I (Hokudo Co. Ltd., Sapporo, Japan) were used as the primary antibody and a goat anti-mouse or anti-rabbit IgG antibody, as the secondary antibody. Horseradish peroxidase (HRP) and hydrogen peroxide were utilized for color development according to the manufacturer`s instructions (Vectastatin kit, Vector Lab., CA).

Clinical assessment was performed using RECIST criteria [23] as follows; complete response (CR) : disappearance of lesions at 4 wks, partial response (PR): 30% decrease in sums of longest diameters at 4 wks, stable disease (SD): neither PR nor PD criteria met, progressive disease (PD): 20% increase in sums of longest diameters. Clinical response was rated as maximal through the DC vaccinations. The patients received up to 10 injections on the condition that at least one measurable lesion showed more than a SD response and/or an ELISPOT assay performed after 4 injections indicated a positive response for more than 1 peptide. Adverse effects were evaluated according to the NCI Common toxicity criteria after 4 DC injections. Peripheral blood mononuclear cell (PBMC) samples were harvested before and 29, 78, 134 and 190 days after the start of DC injections for immunological monitoring. All patients were followed regularly throughout, and an MRI was performed every 2 to 3 months during the vaccination period.

The ELISPOT assay was performed using PBMCs drawn prior to vaccination and after 4 DC injections. Briefly, PBMCs were incubated in a 24-well culture plate and divided into non-adherent and adherent cells. Adherent cells were treated with a peptide cocktail and β2-microglobulin, and co-cultured with non-adherent cells in the presence of IL-2 and IL-7. On day7, non-adherent cells were re-stimulated with peptide-pulsed adherent cells. On day14, responder cells (5×10⁴/well) were stimulated with HLA-A2 or A24 peptides in a 96-well culture plate coated with anti-IFN-γ antibody (MABTECH AB, Nacka, Sweden) overnight. Finally, positive spots stained with anti-IFN-γ antibody were measured using the KS ELISPOT system (Carl Zeiss AG, Overkochen, Germany). A HLA-A2- (GILGFVFTL) or A24-restricted CMVpp65 peptide (RYSIFFDY) was used as a positive control. The spot number per well of peptide-stimulated CTLs was compared to that of a negative well without peptide using Student’s paired two-tailed t-test.

PBMCs were stimulated with 25 ng/ml of PMA (Sigma) and 1 μg/ml of ionomycin (Sigma) for 5hrs in a 96-well culture plate. After the stimulation, cells were stained with FITC-anti-CD4 MoAb, and subsequently intracellular staining was performed with fix/permealization buffer and PE-labeled anti-IFN-γ or anti-IL-4 MoAb (Pharmingen,San Diego, CA). Finally, the ratio of Th1 (IFN-γ⁺) and Th2 (IL-4⁺) was calculated in PBMC samples obtained from patients.

The HLA-A2 or A24 peptide solution and KLH at a dose of 50 μg/ml were injected intradermally into the forearm and the redness and induration at the injection site were measured on days 29, 78, 134 and 190 after the 1st DC injection. 1 × 10⁶ DCs treated with peptides were added to DTH antigens after the start of the vaccination. PPD was used as a positive control.

Statistical differences were analyzed using Student’s paired two-tailed t-test. Values of p < 0.05 were considered significant.

An analysis of tumor antigen expression by RT-PCR and IHC was performed in 6 evaluable cases. The antigen expression was determined as positive, when either the RT-PCR or IHC analysis was positive. All 5 tumor antigens including MAGE-A1, -A3, HER2, gp100 and WT1 were positive in 5 cases except for patient 5 in which 3 antigens were identified in the tumor (Table 2). A representative case of tumor expression analyzed by IHC, patient 6, showed strong reactions to MAGE-A1, MAGE-A3, and WT-1 antigens (Figure 1). The HLA-class I protein was positive in all cases, but the expression level was different (data not shown).

Through gradient-centrifugation using Optiprep, the frequency of CD14⁺ cells improved from 11.9% to 44.6%, an increase of 4 fold (Table 3). The mean percentage of DCs rated as lin^-HLA-DR⁺ (Figure 2) and the DC1/DC2 ratio in glioma patients was 56.9 ± 19.1% and 35.6, respectively, comparable to a previous phase I DC vaccine study for metastatic melanomas. Comparing obtained DC numbers to starting total cell numbers, the recovery rate was determined as 8.79%, also comparable to the previous study. Importantly, IL-12p70 levels produced by DCs through CD40 ligand stimulation were remarkably high, more than 1,000 pg/ml, which indicated a fully mature phenotype of the obtained DCs (Table 3). The frequencies of the DC-related markers were determined in gated lin^-HLA-DR⁺ cells as follows; CD1a 16.3 ± 13.2%, CD40 95.9 ± 5.1%, CD54 99.3 ± 1.2% , CD80 92.4 ± 15.5%, CD83 44.7 ± 19.5%, CD86 99.6 ± 0.3%, CD205 73.9 ± 12.1%, CD207 92.3 ± 11.6%, HLA-class I 99.0 ± 1.2%, which also demonstrated the fully mature phenotype (Table 4).

CTL precursors against synthetic peptides were recognized after the vaccination in 6 evaluable cases (Table 5). Notably, 3 cases demonstrated CTL responses to more than 3 peptides, and patient 2 showed a durable SD without relapse for 34 months after the start of the vaccination, with a remarkably high CTL response to 4 HLA-A24 peptides (Figure 3).

In all 9 cases, the balance of Th1 and Th2 shifted more to Th1 (mean ratio; 13.1 ± 7.3) before DC injections (Table 5). Unfortunately, PBMC samples were not obtained after the vaccinations in 8 cases. No correlation of Th1/Th2 ratio to ELISPOT responses was observed.

Positive DTH tests were verified in 4 patients against all peptides, 4 against DCs treated with peptides, and 4 against KLH (Table 5). Specifically, 2 patients (2 and 9) tested positive for all 3 antigens, and patient 2 exhibited very strong reactions. Surprisingly in patient 2, after the start of the vaccination, the responses to each peptide, KLH and DCs increased to a plateau, and responses to KLH and DCs remained highly positive even after more than 2 years, despite that the response to peptides went down after the vaccination ceased (Figure 4).

Safety was assessed after 4 DC injections in all 9 cases. Mild hepatic dysfunction (grade II) was seen in 1 case, however it was only temporary and disappeared in spite of the continuance of DC injections. No clinical symptoms of autoimmune disease were found (Table 5).

Clinical response was rated as maximal throughout the DC vaccinations. Seven of 8 showed PD due to a rapid progression of the disease (Table 5). Patient 2 demonstrated a long SD after total resection for more than 2 years, in which strong CTL and DTH responses were recognized throughout the vaccination period.