Abstract B036: Systemic and tumor-microenvironment inflammation shape outcomes in patients with immunologically cold, treatment-refractory tumors treated with Fc-enhanced anti–CTLA-4 botensilimab

Abstract Botensilimab (BOT), a novel Fc-enhanced anti–CTLA-4 antibody, alone or combined with balstilimab (BAL; anti–PD-1), has demonstrated promising clinical activity across immunologically “cold” treatment-refractory cancers. However, conventional biomarkers, such as tumor mutational burden and PD-L1, have limited ability to predict patient response, hindering optimal patient selection. We hypothesized that integrating systemic inflammatory markers with tumor microenvironment (TME) and peripheral immune features could better elucidate clinical outcomes for BOT±BAL therapy. We assessed systemic, tumor, and clinical features of patients treated with BOT±BAL for metastatic treatment-refractory cancers in the phase 1b C-800-01 trial (NCT03860272). Patients had cold cancers (e.g., microsatellite-stable colorectal, sarcoma, and ovarian cancer) or were relapsed/refractory (R/R) to conventional anti–PD-(L)1/CTLA-4 therapies (hepatocellular carcinoma and non-small cell lung cancer). Multi-modal baseline biomarker assays included plasma cytokines, peripheral blood mononuclear cell flow cytometry, tumor immunohistochemistry, bulk RNA-seq, and whole-exome sequencing. We analyzed exploratory efficacy associations. Hazard ratios (HR) were analyzed by Cox proportional-hazards models. Complementary machine-learning feature-selection methods were used. As of October 2025, 384 patients receiving ≥1 mg/kg BOT±BAL (BOT alone, n=43) were efficacy evaluable with biomarker data; median overall survival (mOS) was 16.4 mo. Shorter mOS occurred with high baseline systemic inflammation markers, including IL-6 (10.3 mo, HR=1.94, P<.001) and C-reactive protein (12.1 mo, HR=1.50, P<.001), and with high alkaline phosphatase (12.0 mo, HR=1.37, P<.001). Longer mOS was associated with high TME baseline immune activation signatures (T cell-inflamed: 31.3 mo, HR=0.42, P=.0001; IFN-γ-responsive genes: 20.9 mo, HR=0.42, P=.0001), intact TME antigen-presentation machinery (18.5 mo, HR=0.35, P=.0016), and more differentiated peripheral effector memory CD4 T cell states (20.9 mo, HR=0.72, P=.0188). Patients R/R to prior anti–PD-(L)1/CTLA-4 showed higher baseline immune activation but also showed TME resistance markers (high T regulatory [Treg] to non-Treg ratio, elevated exhaustion-signature gene expression, and dampened on-treatment BOT±BAL responses) compared to immunotherapy-naïve patients. Immune-mediated adverse events within the first 12 weeks of treatment correlated with longer mOS, baseline CD4/CD8 T cell ratios, and on-treatment IL-17/rhodopsin-like receptor pathway activation. Additional analyses are ongoing. In cold and treatment-refractory cancers, clinical outcomes with BOT±BAL appear to depend on the balance between detrimental systemic inflammation and beneficial TME immune activity plus peripheral T cell differentiation, rather than conventional single biomarkers. These findings suggest that integrating peripheral and TME markers across biological pathways may enable reliable patient stratification for next-generation immunotherapy. Citation Format: Chloe Delepine, Sean Bourdelais, Gabriel Mednick, Bishnu Joshi, Justin G. Keith, Erin Driscoll, Alexis Rozen, Eli Fellus, Wei Wu, Joseph E. Grossman, Benny Johnson, Steven J. O'Day, Dhan Chand. Systemic and tumor-microenvironment inflammation shape outcomes in patients with immunologically cold, treatment-refractory tumors treated with Fc-enhanced anti–CTLA-4 botensilimab [abstract]. In: Proceedings of the AACR Immuno-Oncology Conference (AACR IO): Discovery and Innovation in Cancer Immunology: Revolutionizing Treatment through Immunotherapy; 2026 Feb 18-21; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Immunol Res 2026;14(2 Suppl):Abstract nr B036.