Background: 1-yr OS is ~40% for HER2- & ~55% for HER2+ advanced (aGEA). Targeted therapies (tx) have had limited benefit due to molecular heterogeneity. Methods: This phase 2a study of a personalized tx strategy (PTS) enrolled newly diagnosed aGEA pts who then received up to 3 cytotoxic (cx) lines: first line (1L) 5FU + oxaliplatin, 2L 5FU + irinotecan & 3L 5FU + docetaxel. Baseline biomarker profiling (BP) was mandated on primary & metastatic tumors (PT/MT) & progressive disease points (PD1, PD2). Assigned antibody (AN) was added to cx by a predefined prioritized tx algorithm (PTA) (Table) based on the MT BP. At PD1, pts went to 2L cx + initial AN. Upon results of PD1 BP, pts changed AN only if BP evolved per PTA. The same was done at PD2. If AN was unavailable (MET/FGFR2), these pts were tx’d with cx alone (not ITT). The 1⁰ endpt was 1-yr OS of the PTS. Assuming historical 50% 1-yr OS for all aGEA pts, 68 pts tx’d per protocol PTS provided 80% power to detect an HR=0.67, corresponding to a 1-yr OS rate of 63% (under exponential survival), using a 1-sided test at the 0.10 alpha level. 2⁰ endpts: safety, feasibility, PT/MT BP discordance at baseline & over tx line, & OS/PFS/ORR by tx line & BP group. Results: Between 6/2015-5/2019, 80 consecutive pts enrolled at 3 sites: ECOG PS 0-2 40/33/7; Male 80%; median age 60, range 28-81, peritoneal disease 36%. AN assigned by PTA at 1L & 1-yr OS are shown (Table). PT/MT discordance was 37%. Of 68 pts treated by PTS ITT, the 1-yr OS was 69.4% (p<0.001). The mOS was 16.4m [95%CI 13.8-20.8]. Any grade >3 tox thru all 3 tx lines was seen in 32% of pts. 2⁰ analyses will be presented. Conclusions: PANGEA was feasible & met its 1⁰ efficacy objective with observed 1-yr OS of 69.4%, meriting a randomized study. Clinical trial information: NCT02213289. U.S. National Institutes of Health.

1 MT priority over PT

2 priority over HER2 only 2L+

3 priority to higher gene copy

4 excluded from ITT 5 evaluable