Background: It is generally assumed that adherence to clinical practice guidelines can reduce treatment variability and improve patient outcomes. Gastric cancer clinical treatment pathways were implemented in the US Oncology Network (USON) in accordance with the National Comprehensive Cancer Network (NCCN) gastric cancer guidelines and refined since 2010. This study was designed to evaluate the impact of 4 pathway periods (PP): pre-pathway: pre-Aug 2010; Level 1: Sept 2010-Nov 2014; Clear Value, Dec 2014-Feb 2017; Value: post-Mar 2017 with respect to concordance with NCCN treatment recommendations, treatment heterogeneity and outcomes, including time to treatment failure (TTF) for both for first-line (1L) and second-line (2L) therapy. Methods: Adult patients in the USON iKnowMed electronic medical records system were eligible if they were treated at a participating site and diagnosed with gastric cancer and treated during the study period (2009-2018) to allow for follow up through March 31, 2019. Concordance with NCCN treatment recommendations was assessed by PP based on guidelines that were in effect during the PP and reported as percentages for both 1L and 2L. Treatment heterogeneity was calculated using the Herfindahl-Hirschman Index (HHI), which is evaluated from 0.0 to 1.0 (complete heterogeneity to complete homogeneity), with change of 0.1 being considered meaningful. TTF was defined as time from initiation of the line of therapy until the start of the subsequent line of therapy or death and estimated using Kaplan-Meier method. Results: Of 3191 eligible patients, 2297 had advanced/metastatic disease. Before pathway initiation NCCN concordance was initially low at 25.2% increasing to 66.0% in the most recent PP in the 1L setting and increased from 34.7% during the first PP to 66.7% in the 2L setting. Heterogeneity was initially 0.0853 and 0.0555 and became more homogeneous at 0.2145 and 0.1828 in 1L and 2L settings, respectively. TTF and HHI are presented in Figure 1. Conclusions: The evolution of the USON Pathways program demonstrated higher concordance with NCCN guidelines in 1L and 2L, meaningful decrease in heterogeneity in both 1L and 2L, and increased TTF over time in the 1L setting. This suggests that utilization of guidelines and pathways may improve GC outcomes.