Objective: Biomarker testing is recommended for advanced gastric cancer/gastric esophageal junction cancer (aGC/GEJC) patients to inform treatment decisions. This study aimed to examine biomarker testing patterns among aGC/GEJC patients in U.S community oncology practices.

Methods: Structured data within the iKnowMed electronic health record database of The US Oncology Network were analyzed to identify adult patients diagnosed with aGC/GEJC between 1/2015 to 3/2020. Timing and testing frequency of HER2, PD-L1, MSI, MMR, TMB and NTRK was assessed through 3/2021.

Results: 2398 (aGC=1727, aGEJC=671) patients were identified [mean age 66years; median follow-up 7.9months]. Of these, 69.6% (n=1669) received ≥1 line of therapy and 60.4% (n=1449) received ≥1 biomarker test. Among patients with ≥1 biomarker test, 92.5% were tested for HER2 (n=1340), 38.3% (n=555) were tested for at least one immunotherapy (I/O) targeted biomarker [MSI (26.2%, n=380), PD-L1 (24.6%, n=356), MMR (24.3%, n=352), TMB (0.3%, n=4)], and 5.0% (n=73) for NTRK. Among patients tested for HER2, 24.6% (n=329) had their first HER2 test during earlier stages of GC/GEJC or before aGC/GEJC diagnosis, and 54.5% (n=730) between aGC/GEJC diagnosis and 1L initiation. 383 patients had first HER2 tests before 1L initiation and received subsequent 2L, but of these 148 (38.6%) were not re-tested before initiating 2L. In contrast, PD-L1 testing was more common after 1L initiation: 36.2% (129/356) patients had their first PD-L1 test after 1L and before 2L initiation, and 38.2% (136/356) after 2L initiation [similar distributions were seen for other I/O targeted biomarkers].

Conclusion: About 2 in 5 aGC/GEJC patients didn’t have a record of biomarker testing in structured EHR data. HER2 testing was the most common test before initiation of 1L, however, not all patients requiring 2L were re-tested for HER2. As more targeted therapies become available, future studies are needed to identify barriers to early biomarker testing for aGC/GEJC patients.