Background: The phase 2a, dose-finding JOSHUA study reported increased P clearance (CL; 37% lower P steady-state C_min [C_min,ss]) in pts with HER2+ mGEJC/GC vs metastatic breast cancer (MBC). Based on these data, 840 mg q3w was selected for testing in the phase 3 JACOB study (NCT01774786) to achieve similar P concentrations (conc) to BC studies with the 840 mg then 420 mg q3w P dose. In JACOB, while there was evidence of treatment activity, addition of P to trastuzumab (H) and chemotherapy (CT) in 1^st-line therapy did not significantly improve overall survival (OS) vs placebo (Pla)+H+CT (hazard ratio: 0.84 [95% CI 0.71–1.00], median OS 17.5 vs 14.2 months), in pts with HER2+ mGEJC/GC. Here we report the PK and ER analysis of P in JACOB. Methods: PK samples were collected at Cycles 1–4, 6, and 8 (predose); Cycles 1, 2, 4 and 8 (postdose); and at follow-up. P+H peak and trough conc were summarized by descriptive statistics and % pts with a C_min,ss≥20 μg/mL (target conc) tabulated. PK exposure was compared across geographic regions. The Kaplan-Meier method and a log-rank test were used to assess OS across C_min quartiles. PK drug-drug interactions (DDIs) were assessed using serum C_min geometric mean ratios. Results: Pts with ≥1 dose of P or H and ≥1 PK sample were included (P: n = 374 [P+H+CT]; H: n = 372 [P+H+CT] and 375 [Pla+H+CT]). Mean C_min,ss Cycle 6 for P was 114±51.8 μg/mL, 99.3% of pts had C_min,ss for P ≥20 μg/mL. There were no differences in OS across Cycle 1 and Cycle 6 (s-s) P C_min quartiles (Q1: 0.075–30.6, Q2: 30.6–39.8, Q3: 39.8–51.9, Q4: 51.9–318; n = 349, P = 0.52 and Q1: 0.075–77.1, Q2: 77.1–110, Q3: 110–145, Q4: 145–450; n = 274, P = 0.78, respectively). Mean C_min,ss was comparable across geographic regions. No apparent DDIs were observed for CT on P C_min,ss, or P on H. Two of 349 (0.6%) evaluable pts tested positive for anti-drug antibodies to P; there was no impact on P PK. Conclusions: In JACOB, the 840 mg q3w P dose achieved target C_min,ss in > 99% pts with HER2+ mGEJC/GC. ER analysis showed no correlation between P trough conc and OS. PK data were consistent with prior GC (P: 840 mg) and BC (P: 840 mg/420 mg) studies and higher P CL in HER2+ mGEJC/GC vs other tumor types. Clinical trial information: NCT01774786