Background: ICIs changed the way NSCLC is treated, but not all patients benefit from it. PD-L1 level is used to predict response to therapy, but its performance is sub-optimal. KRAS is important in NSCLC tumorigenesis, but the impact of its mutations in patients treated with ICIs is unclear. Similarly, studies evaluating co-mutations in TP53, STK11 and KEAP1 as well as the NLR showed that they may predict the benefit of ICIs. Methods: We conducted a retrospective study including all consenting patients with NSCLC treated with ICIs at the CHUM between July 2015 and June 2020. OS and PFS were compared in co-mutation subgroups using Kaplan-Meier and logrank methods. Co-mutations in TP53, STK11 and KEAP1 as well as the NLR were accounted for. Overall response rate (ORR) and safety data was also compared in subgroups and will be detailed at the meeting. Results: We included 100 patients with known KRAS status. From these, 50 were wild-type (KRAS^WT) and 50 were mutated (KRAS^Mut). The most frequent mutation was G12C (54%). Co-mutation status for TP53, STK11 and KEAP1 were known for, respectively, 40, 39 and 38 patients. Co-mutations for these genes were present in respectively 19 (47.5%), 8 (20.5%) and 4 (10.5%). Data comparing KRAS^Mut and KRAS^WT showed non-significant differences in survival (median OS of respectively 21.1 vs. 17.7 months, p = 0.27). The presence of STK11 and/or KEAP1 mutations was associated with a negative impact on survival when compared with wild-type (median OS 7.4 vs 20.4 months, p = 0.001). When the presence of a KRAS mutation was compounded with STK11 and KEAP1, KRAS^Mut (vs KRAS^WT) trended to a better prognosis in STK11+KEAP1^WT tumors (median OS of 21.1 for KRAS^Mut vs 15.8 for KRAS^WT, p = 0.15), but not in STK11+/-KEAP1^Mut tumors (7.4 for KRAS^Mut vs 7.0 for KRAS^WT). No influence on survival was seen in relationship to the TP53 co-mutation. Interestingly, the NLR was significantly higher with STK11 mutations (6.66^Mut vs 3.59^WT, p = 00012), slightly lower with TP53 mutations (3.23^Mut vs 4.82^WT, p = 0.047) but not impacted by KEAP1 (3.72^Mut vs 4.20^WT, p = 0.72) or KRAS mutations (4.32^Mut vs 5.21^WT, p = 0.34). Conclusions: The STK11 and KEAP1 mutations are significant adverse predictors of ICI therapy benefit. The NLR is strongly impacted by STK11 mutations but not by KEAP1 mutations suggesting marked differences in the resistance mechanism for both mutations. In STK11-KEAP1^WT tumors, KRAS mutations seems to be associated with improved survival in NSCLC patient treated with ICIs.