associated with progression (PD) and poor outcomes. We reported that pretreatment dNLR ((neutrophils/[leucocytes-neutrophils]; high>3) correlated with immune checkpoint inhibitor (ICI) outcomes in advanced (a) NSCLC pts. Although neutrophil population is heterogeneous, the immature neutrophils (i.e. CD15+CD244-CD16 low , among others) seem to be a key subpopulation linked to PD. Tumor-associated neutrophils (TAN) can be also modulator on the microenvironment. We aimed to assess the role of pretreatment circulating immature-neutrophils and tissue-TAN, combined with dNLR, on ICI outcomes in aNSCLC pts. Method: aNSCLC pts treated with ICI at our institution between 11/2012 and 08/2018 were eligible. Pretreatment immunophenotyping of monocytes, monocytic MDSC (mMDSC) and granulocytes (CD15, CD11b, CD33, CD244, CD16, CD14, CD32, CD64, HLA-DR) was prospectively performed by flow cytometry in fresh whole blood in 58 pts; we defined immature-neutrophils as CD15+CD244-CD16 low. TAN in the stroma were assessed using H&E staining from archival specimen, available from 80 pts. dNLR was retrospectively collected; available from 343 pts. Correlation between baseline circulating neutrophils phenotype, TAN and dNLR was evaluated as well as their impact on outcomes: progression-free survival (PFS), overall (OS), including death before 12 weeks (12wk-death) (fast-PD) Result: 366 pts included; 320 (90%) smokers, median age 63; 280 (77%) nonsquamous, 117 (64%) 1%PDL1 and 183 missing. Median PFS (mPFS) was 1.93 months (m) [95%CI, 1.8-2.3] and mOS 8.8m [6.5-11.6]. Overall, 12wk-death rate was 31% [25.9-35.6]. Pretreatment high-dNLR (143/343; 42%) was correlated with poor PFS (P¼0.002), OS P¼0.0003) and a 12wk-death rate of 43% [34.5-50.9]. Pretreatment high immature-neutrophils (30/58; 53%), defined by logrank maximization method (>0.22%), were also associated with poor PFS (P¼0.04), OS (P¼0.0007) and a 12wk-death rate of 48.7% [26.7-64.1]. TAN (9/80; 11%) were not correlated with outcomes. There was not a correlation between immature-neutrophils, tissue-TAN and dNLR. When evaluating pretreatment immature-neutrophils and dNLR together, we identified a fast-PD phenotype (high immature-neutrophils/high-dNLR, 10/58; 17%), with a mOS of 1.3m [0.73-not reached (NR)] and 12wk-death rate of 60% [14.5-81.3] compared to a responder-phenotype (low immature-neutrophils/low-dNLR, 12/58; 21%), associated with good outcomes: mOS NR [18.23-NR] (P¼0.002). Conclusion: Pretreatment high circulating immature-neutrophils (CD15+CD244-CD16 low) correlate with early failure to ICI and fast-PD phenotype. The combination of circulating immature-neutrophils and dNLR could improve the identification of this population. The impact of immature-neutrophils on ICI should be more deeply explored.