BACKGROUND: Patients with relapsed CNS malignancies or DIPG face terrible prognoses. We hypothesized that T cells specific for 3 tumor-associated antigens (TAA), WT1, PRAME and survivin, would be safe and elicit anti-tumor immunity. Methods: Patients (n=15) received autologous tumor antigen-associated T cells (TAAT) (up to 4x107/m2) for newly diagnosed DIPG (Group A) or recurrent CNS malignancies (Group B) on a Phase I dose-escalation study (NCT03652545) and were monitored for safety and response. RESULTS/ Discussion: 15/15 patients who received TAAT completed the 45-day safety monitoring phase with no dose-limiting toxicities. Adverse events were minimal despite multiple pretreatments in Group B. Infused cells were predominantly CD3+ T cells (median 96%; range: 87-99%), with CD4+ and CD8+ comprising 16% (range: 5-87%) and 40% (range: 4-67%) respectively. Specificity for 1-3 TAAs was demonstrated in 13/15 TAAT by a-IFN-γ ELISPOT. Plasma cytokine and proteomic analyses are ongoing but have demonstrated dynamic post-infusion immune cytokine and protein responses. Increases in the inflammatory and immune-stimulatory cytokines IL-1b, IL-6, IL-2 and IL-7 were observed post-infusion in most patients evaluated. Infusion-related increases in regulatory cytokines IL-10 and IL-13 were also observed in 4/7 patients. These results are consistent with an infusion-mediated immune response in vivo. Of 9 patients who have been tested thus far, 29/92 plasma proteins showed significant differences between dose levels 1 and 2, including increased IL-7 (p < 0.0004) and CD40L (p < 0.046) and reduced IL-4 (p < 0.0004). T cell receptor sequencing data on in vivo TAAT persistence is pending. In summary, TAAT have thus far been safe and elicit immune responses in vivo. Clinical and immunologic response assessments are ongoing.