Background: In a prior analysis, we identified 525 pts with newly diagnosed NSCLC 4 harboring AOD in the ICD. Of these, 141 were treated prior to the reporting of AOD with chemotherapy (C), immune checkpoint inhibitor (ICI), or both. This report details the clinical outcomes of these 141 compared to the 384 treated after AOD reported. Methods: Real world data (RWD) were obtained from a curated ICD for pts with NSCLC 4 diagnosed 1/1/2018-12/31/2020 with cutoff of data 3/31/2021. Pts with EGFR, ALK, ROS1, BRAF, MET, RET, HER2, and NTRK were included if their treatment record were captured. Also included were demographics, ECOG score, date of first report of AOD and dates of initiations of first and any second line of therapy and date of apparent death (AD). Outcome measures were time to next treatment or apparent death (TTNT) and apparent survival (AS) (ICD model does not allow date of death per HIPAA de-identified Expert Determination). Descriptive statistics were used with Kaplan-Meier (K-M) estimates and Hazard Ratios (HR) by Cox regression. 3 cohorts were defined: Group (Gr) A with 384 pts treated after AOD reported and used as the comparator; the 141 pts treated before AOD with C, ICI or both were divided into Gr B (n = 51) who subsequently switched to TKI within 35 days and Gr C (n = 90) who did not switch. Results: As shown in data table, AS was significantly worse in Gr B and Gr C, TTNT was significantly worse in Gr C and with worsening trend in Gr B. Two potential confounders were identified: higher ECOG scores might indicate more urgency to assign treatment, but pts with ECOG ≥ 2 were similar in all 3 groups; also, difference in proportion of EGFRm by Group (Gr A 62%, Gr B 57%, and Gr C 29%), but separating cohorts by EGFR mutation status did not alter results. Conclusions: While subject to the limitations of a retrospective observational RWD study, this study strongly suggests outcomes are significantly compromised in pts harboring AOD but who are treated initially with C, ICI or both, even in pts quickly switched to TKI. Since a prospective clinical trial is not ethically feasible, these findings may stimulate review of current guidelines, fuel optimization of universal testing in NSCLC 4, and encourage utilization of liquid or ultra-fast NGS with their rapid turnaround times in order to improve survival in this setting.