Generation of polyclonal CMV-specific T cells for the adoptive immunotherapy of glioblastoma

Abstract

Glioblastoma (GBM) is the most common primary brain cancer in adults and is virtually incurable. Recent studies have shown that cytomegalovirus (CMV) is present in majority of GBMs. To evaluate whether the CMV antigens pp65 and IE1, which are expressed in GBMs, could be targeted by CMV-specific T cells, we measured the frequency of T cells targeting pp65 and IE1 in the peripheral blood of a cohort of 11 sequentially diagnosed CMV-seropositive GBM patients, and evaluated whether it was feasible to expand autologous CMV-specific T cells for future clinical studies. All 11 CMV-seropositive GBM patients had T cells specific for pp65 and IE1 in their peripheral blood assessed by IFNγ enzyme-linked immunospot assay. However, the precursor frequency of pp65-specific T cells was decreased in comparison with healthy donors (P=0.001). We successfully reactivated and expanded CMV-specific T cells from 6 out of 6 GBM patients using antigen-presenting cells transduced with an adenoviral vector encoding pp65 and IE1. CMV-specific T-cell lines contained CD4 as well as CD8 T cells, recognized pp65 and IE1 targets and killed CMV-infected autologous GBM cells. Infusion of such CMV-specific T-cell lines may extend the benefits of T-cell therapy to patients with CMV GBMs.

Figure 1. Immunohistochemistry (IHC) for pp65 and IE1 and CMV genome-specific in situ hybridization (ISH) in primary GBM samples

A. IHC was performed on 11 parraffin embedded primary GBM samples. Overall, 5 of 11 samples showed positivity for pp65 while 10 out of 11 patient samples showed positivity for IE1 by IHC. B. Eight of 9 samples were positive using ISH. For both tests, CMV-infected lung tissue was used as a positive control. Elimination of the primary antibody and an ALu tandem probe were used as negative controls for IHC and ISH, respectively.

Figure 2. Cellular immune response against CMV in CMV-seropositive GBM patients

A. Precursor frequency of pp65- and IE1-specific T cells in the peripheral blood of CMV-seropositive GBM patients at diagnosis compared to 7 healthy CMV-seropositive controls. B. Heatmap summarizing the results of the tissue reactivity (IHC and ISH) as well as the corresponding pp65 and IE1 precursor frequency in the peripheral blood (PB) of our patient cohort. SD: standard deviation from the mean value for healthy donors.

Figure 3. Expansion and phenotype of CMV-specific T-cell lines

A. CMV-specific T-cell lines from 6 patients were expanded to clinically relevant levels within 6 weeks of culture. B. Phenotype of 6 T-cell lines showing predominance of CD8-positive TCRαβ T cells.

Figure 4. Functionality of CMV-specific T-cell lines

A. IFNγ ELIspot of CMV-specific T cells generated from CMV-seropositive patients. All 6 T-cell lines contained T cells specific for pp65 and IE1. B. Pentamer analysis was performed on 4 representative cell lines confirming the presence of pp65-specific T cells.

Figure 5. Polyfunctionality of CMV-specific T-cell lines

Upon exposure to pp65 and IE1 pepmixes, 35–42% of CMV-specific T-cell lines co-secreted IFNγ and TNFα indicating polyfunctionality. A representative dot plot is shown for the intracellular cytokine analysis using flowcytometry of CMV-specific CTL line from patient GBM7 and GBM8 and the pie chart summarizes the mean values derived from these two lines. Note that a subset of responding T cells co-secretes both cytokines.

Figure 6. CMV-specific T-cell lines recognize and kill CMV-expressing autologous GBM cells

⁵¹Chromium-release assay of representative CMV-specific CTLs against autologous (A.) pp65 and IE1 peptides pulsed and (B.) autologous CMV-infected GBMs cells. CMV-specific T-cell lines from representative patients (GBM1: HLA-A29,74, -B18,44; GBM2: HLA-A31,68, -B27,52) recognized and killed CMV-infected as well as pp65- and IE1-pulsed autologous glioma cells vs. control. No killing of the HLA-mismatched GBM cell line U373 (HLA-A2, -B18) after infection with CMV VR1814 was observed.