An optimized method for manufacturing a clinical scale dendritic cell-based vaccine for the treatment of glioblastoma

Abstract

Immune-based treatments represent a promising new class of therapy designed to boost the immune system to specifically eradicate malignant cells. Immunotherapy may generate specific anti-tumor immune responses, and dendritic cells (DC), professional antigen-presenting cells, are widely used in experimental cancer immunotherapy. Several reports describe methods for the generation of mature, antigen-pulsed DC for clinical use. Improved quality and standardization are desirable to obtain GMP-compliant protocols. In this study we describe the generation of DC from 31 Glioblastoma (GB) patients starting from their monocytes isolated by immunomagnetic CD14 selection using the CliniMACS® device. Upon differentiation of CD14+ with IL-4 and GM-CSF, DC were induced to maturation with TNF-α, PGE(2), IL-1β, and IL-6. Whole tumor lysate was obtained, for the first time, in a closed system using the semi-automated dissociator GentleMACS®. The yield of proteins improved by 130% compared to the manual dissociation method. Interestingly the Mean Fluorescence Intensity for CD83 increased significantly in DC pulsed with "new method" lysate compared to DC pulsed with "classical method" lysate. Our results indicate that immunomagnetic isolation of CD14(+) monocytes using the CliniMACS® device and their pulsing with whole tumor lysate proteins is a suitable method for clinical-scale generation of high quality, functional DC under GMP-grade conditions.

Figure 1. Lysate protocol.

The modified lysate protocol (New method, black bar, 19 productions) increased the yield of protein obtained per gram of tissue by 130% with respect to the Old “Classical method” (empty bar, 12 productions). Data are represented as mean mg of proteins per gram of tissue. Statistical bars on graph indicate the SD value.

Figure 2. Phenotypic analysis of DC. a)

Phenotypic analysis of maturation-associated markers was determined by flow cytometry on iDC, aDC and on mDC. Variation of the percentage of positive cells expressing maturation markers (CD80, CD83, CD86) and down-regulation of CD14 (monocytes marker) are represented. A statistically significant difference for maturation markers was present between mDC (black bar) and both iDC (gray bar) and aDC (empty bar) (p<0.05); no differences were observed in HLA-DR expression. Statistical bars on graph indicate the SD value. The overlay representation of histograms illustrates the up regulation of CD80, CD83, CD86 and the down regulation of CD14 during culture. Only a slight up-regulation was observed on HLA-DR expression during maturation stages (p>0.05). Histograms are representative of one of 31 independent productions analysed. iDC (red line); aDC (green line); mDC (blue line); monocytes (black line). b) CD83 expression resulted higher and more homogeneous in DC activated with tumor lysate produced following the “new method” (black bar, 19 productions) than in DC activated with tumor lysate produced following the old “classical method” (empty bar, 12 productions). Data are expressed as MFI normalized on isotype control. The overlay representation of histograms reveals that CD83 expression resulted more homogeneous in the second group of DC (“new method”, purple line) respect to the first one (old “classical method”, yellow filled).

Figure 3. Functional evaluation of DC.

The ability of iDC versus mDC in the antigen presentation was compared evaluating the potency of DC as their in vitro allo-stimulatory capacity of PBMC from healthy volunteers. Final product (mDC, black bar, 31 experiments) resulted more potent than its immature counterpart (iDC, empty bar, 18 experiments) in the MLR induction. We compared new lysate activated and old lysate activated cells for their ability in the induction of allogeneic MLR functional assay. “New method-tumor lysate” (19 productions) activation seem to slightly improve the functional ability of mDC in their allo-stimulatory ability respect to “classical method-tumor lysate” (12 productions) activation but this difference is not statistically significant. By contrast no proliferative responses were induced by antigen-loaded mDC in the lymphocyte of GB patients prior to vaccination (red bar, 18 experiments). Data are expressed as Stimulation Index (SI); statistical bars on graph indicate the SD value.