Glioblastoma-mediated Immune Dysfunction Limits CMV-specific T Cells and Therapeutic Responses: Results from a Phase I/II Trial.

Cytomegalovirus (CMV) antigens occur in glioblastoma but not in normal brains, making them desirable immunologic targets. Highly functional autologous polyclonal CMV pp65-specific T cells from patients with glioblastoma were numerically expanded under good manufacturing practice compliant conditions and administered after 3 weeks of lymphodepleting dose-dense temozolomide (100 mg/m2) treatment. The phase I component used a 3+3 design, ascending through four dose levels (5 × 106-1 × 108 cells). Treatment occurred every 6 weeks for four cycles. In vivo persistence and effector function of CMV-specific T cells was determined by dextramer staining and multiparameter flow cytometry in serially sampled peripheral blood and in the tumor microenvironment. We screened 65 patients; 41 were seropositive for CMV; 25 underwent leukapheresis; and 20 completed ≥1 cycle. No dose-limiting toxicities were observed. Radiographic response was complete in 1 patient, partial in 2. Median progression-free survival (PFS) time was 1.3 months [95% confidence interval (CI), 0-8.3 months]; 6-month PFS was 19% (95% CI, 7%-52%); and median overall survival time was 12 months (95% CI, 6 months to not reached). Repeated infusions of CMV-T cells paralleled significant increases in circulating CMV+ CD8+ T cells, but cytokine production showing effector activity was suppressed, especially from T cells obtained directly from glioblastomas. Adoptive infusion of CMV-specific T cells after lymphodepletion with dose-dense temozolomide was well tolerated. But apparently CMV seropositivity does not guarantee tumor susceptibility to CMV-specific T cells, suggesting heterogeneity in CMV antigen expression. Moreover, effector function of these T cells was attenuated, indicating a requirement for further T-cell modulation to prevent their dysfunction before conducting large-scale clinical studies.