CTIM-12. DUAL IMMUNE CHECKPOINT BLOCKADE FOR SURGICALLY ACCESSIBLE RECURRENT GLIOBLASTOMA: CLINICAL AND MOLECULAR OUTCOMES

Abstract Our previous study of neoadjuvant PD-1 immune checkpoint blockade (ICB) in surgically-accessible recurrent glioblastoma suggested that single-agent PD-1 blockade induced adaptive immune responses that may have paradoxically inhibited functional anti-tumor immune responsiveness. To address this, we performed a surgical-window-of-opportunity trial in resectable, relapsed glioblastoma patients (n=60). Patients received either dual ICB (Arm 1, n=25) with ipilimumab (3 mg/kg) and nivolumab (240 mg), single-agent PD-1 ICB (Arm 2, n=25) with nivolumab (240 mg/kg), or a placebo infusion (Arm 3, n=10) prior to surgical resection. Following surgery, Arms 1 and 3 received dual ICB, while Arm 2 continued on nivolumab until progression or unacceptable toxicity. We collected blood and tumor tissue from all patients and performed RNA and TCR sequencing, along with multiplexed immunofluorescence, to evaluate intratumoral and systemic immune response. Dual ICB induced greater T-cell/interferon-related gene expressions in tumor compared with PD-1 ICB or a placebo infusion. It also resulted in the highest TIL density, as assessed by TCR sequencing. In a multivariate analysis, TIL density strongly correlated with survival outcomes. In PBMCs, dual ICB induced robust interferon gene signatures, and early upregulation of these signatures immediately following treatment was associated with longer survival. However, alongside interferon signaling, monocyte-related signatures also significantly increased following dual ICB, and their persistently elevated levels during adjuvant treatment cycles were linked to poorer outcomes. The median overall survival of patients who received dual ipilimumab+nivolumab was 405 days (95% CI, 265 to NA), while that of single agent nivolumab was 275 days (95% CI, 166 to NA). Our results demonstrate a clear pharmacodynamic effect of dual ICB in glioblastoma and the survival times compare favorably with similar clinical studies done in the same population. Moreover, our transcriptional analysis underscores the importance of balancing immune activation and chronic inflammation in shaping ICB responsiveness.