Abstract The introduction of checkpoint inhibitors (CPI) has revolutionised the treatment of advanced cancers, which act to ameliorate the anti-tumour response through blocking inhibitive immune checkpoint receptors. Whole-exome sequencing (WES) based estimates of tumour mutational burden (TMB) has been robustly associated with response to CPI therapy across multiple tumour types and a threshold of ≥10 mutations per megabase (Mb) is now an FDA-approved biomarker for treatment with pembrolizumab. However, studies to date have been limited to whole exome sequencing (WES), with no data yet reported on the utility of whole genome sequencing (WGS) as a biomarker to predict CPI response. WGS data can identify non-coding alterations which impact the expression of immune regulatory genes, variants likely to form non-coding epitopes, including novel open reading frame mutations especially structure variants (SV). Here, we have identified a pan-cancer cohort of n=364 patients treated with CPI from Genomics England 100,000 Genomes Project (GEL). We find that SV burden is highly associated with worse CPI response (HR =1.3 (1.0-1.8) , p-value = 0.037) and high SV burden group have low tumour-infiltrating lymphocyte (TIL) score (P-value < 0.0001). However, we found that patients who have high proportion of SV derived neoantigen group improved overall survival (HR = 0.44 (0.27-0.73) , p-value = 0.002). In addition, enhancer region alteration of antigen present gene was observed in 43 patients (12%) and was associated with worse CPI response. (HR = 1.5 (1.0-2.2), P-value = 0.034). We further explored the SV derived neoantigen generate neoantigen-specific T cell response. Summary of Cox Model from structure variant Variables No. patients HR (95% CI) P Sex Female 164 Male 200 1.15 ( 0.87-1.52) 0.3285 Cancer Type Others 59 Lung 88 1.39 (0.91-2.14) 0.1255 Melanoma 155 0.84 (0.56-1.28) 0.4267 Renal 62 0.84 (0.52-1.38) 0.495 TMB estimated WGS 364 0.98 (0.97-0.99) 0.0036 SV burden Low 182 High 182 1.36 (1.02-1.81) 0.0358 Proportion of anti-tumour antigen from SV Low 49 High 47 0.48 (0.29-0.8) 0.0045 antigen non-generate 268 0.49 (0.35-0.7) < 0.0001 Citation Format: Hongui Cha, Benjamin Simpson, Charles Swanton, Kevin Litchfield, Genomics England Research Consortium. Somatic structure variants as a biomarker to predict immune checkpoint inhibitor response by generating anti-tumour antigen [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3137.