Background: cGAS-STING (CGS) is an intracellular DNA sensing pathway that when activated, stimulates interferon and other immune-related signaling pathways. While increased expression of CGS-related genes has been associated with better survival in NSCLC, studies investigating this phenomenon in tumors less likely to respond to ICIs such as PDL1-negative, TMB-Low NSCLC are lacking. Here we report the role of CGS pathway in NSCLC tumors that are both PDL1-negative and TMB-Low. Methods: DNA (592-gene or whole exome) and RNA sequencing (whole transcriptome) was performed for NSCLC specimens (n=7264) submitted to Caris Life Sciences (Phoenix, AZ). PDL1 negativity was defined as <1% TPS score (22C3) & TMB-Low as <10 mutations/Mb. Hierarchical Agglomerative Clustering (HAC) was performed on STING pathway genes (CXCL10, CCL5 and GZB) and tumors were grouped into STING-high (n=3070), STING-intermediate (n=1609) and STING-low (n=2585) groups. Microenvironmental cell populations (MCP)-counter analysis was used to estimate cellular composition of immune cells. Statistical significance was determined using chi-square/Fisher’s exact/Mann-Whitney U tests and adjusted for multiple comparisons (q<0.05). Kaplan Meier method was used to estimate the overall survival (OS) derived from insurance claims of IO treated pts with NSCLC (n=400). Pts were stratified into groups above or below the median expression of the STING gene (TMEM173). Results: Compared to STING-low, tumors that were STING-high had enrichment in immune cell infiltrates including CD8+ T cells (4.4-fold) and B-cells (5.9-fold); Additionally, STING-high cohort had significantly higher expression of immune checkpoint (IC) genes including CD274(2.5-fold), CD80 (2.7-fold, all q<0.05). STING-high tumors also had a higher prevalence of TP53 mutations (mts) and a lower prevalence of STK11 and KEAP1 mts (Table). Median overall survival (OS) was 21.1 months in the STING-high cohort (n=200) vs 11.9 months in the STING-low cohort (n=199) in IO treated pts (HR = 0.62, p<0.000.1) Conclusions: To our knowledge this is the largest real-world dataset indicating that enrichment of both the STING gene, and gene signatures associated with the STING pathway represent valuable transcriptomic biomarkers to stratify pts with PDL1-negative and TMB-Low NSCLC who benefit more from IO therapies. Conversely, novel IO combination strategies need to be employed in STING-low tumors to enhance outcomes.