Abstract 6425: Harnessing big data from pan-cancer concerning more than 1300 patients with immune checkpoint inhibitors identifies novel predictive biomarkers

Abstract

Abstract Immune-checkpoint inhibitors (ICIs) have revolutionized cancer therapy, yet many patients face the challenges of either not benefiting or developing resistance to ICIs. Advances in genome sequencing allow the exploitation of high-dimension oncological data in the research and development of precision immuno-oncology. We conducted a comprehensive literature review involving cancer patients treated with ICIs [anti-programmed cell death protein 1 (PD-1)/programmed death-ligand 1 (PD-L1), anti-cytotoxic T-lymphocyte antigen 4 (CTLA-4), or their combination] and collected available transcriptomic data from pretreatment biopsies along with corresponding clinical outcomes. Biomarkers for ICI response were independently investigated within each study using the logistic regression method and combined with the odds ratio (OR) for meta-analysis. A total of 210 formalin-fixed paraffin-embedded pan-cancer samples with pretreatment biopsies from the National Cancer Center retrospective cohort were collected for validation. Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and immunohistochemistry (IHC) were employed to validate the stable and clinically applicable utility of the genomic biomarker identified in our analysis. Across 18 studies, we identified 1106 cancer patients who received immunotherapy with pre-treatment transcriptomic data, including melanoma, non-small cell lung cancer (NSCLC), renal cell carcinoma (RCC), stomach adenocarcinoma (STAD), glioblastoma (GBM), and bladder urothelial carcinoma (BLCA), of which 8421 genes were available in all datasets. Meta-analysis based on a common effect model revealed that 332 genes were positively correlated with a favorable response to ICIs, and 404 genes were negatively associated with the therapeutic outcome, with all pooled OR being statistically significant (P <0.05). We identified a novel biomarker, ZBED2 (Zinc Finger BED-Type Containing 2), which demonstrated superior predictive value over other biomarkers for ICI efficacy (pooled OR, 0.596; 95% CI, 0.450-0.788, P =0.002). Similarly, RT-qPCR (AUC: 0.821) and IHC (AUC: 0.789) results further validated the predictive utility for ICI response in the real-world cohort of 210 cancer patients with pretreatment samples receiving ICIs. Additionally, we observed that the expression of ZBED2 was favorably associated with the overall survival of patients receiving immunotherapy (HR: 0.72). In summary, our study provides a in-depth analysis of potentially predictable biomarkers for ICIs in cancer and highlights the potential of large-scale meta-analyses in identifying stable biomarkers. ZBED2 emerges as a novel predictive biomarker for ICIs and may guide the implementation of risk-related therapeutic strategies. Citation Format: Peng Wu, Chaoqi Zhang, Ao Shen, Xuanyu Gu, Nan Sun, Jie He. Harnessing big data from pan-cancer concerning more than 1300 patients with immune checkpoint inhibitors identifies novel predictive biomarkers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6425.