Abstract

392 Background: Peritoneal metastases (PM) in gastric cancer (GC) portend a poor prognosis. Yet, there are a paucity of data on underlying genomic alterations predictive of GCPM. Methods: Two-hundred and nine patients who underwent primary tumor (PT) surgical resection were included in the prospective cohort. Whole exome sequencing (WES) of these resected PT GC specimens (average coverage 129X for tumor samples, 70X coverage for matched normal/blood controls) were undertaken. Participants were followed-up longitudinally for peritoneal metastases. Results: Fifty-one (51/209, 24.4%) patients developed PM after a median follow-up duration of 87.0 months and had poorer survival outcomes compared to patients without metastasis (log-rank, p<0.001). Patients with bulkier and more advanced tumors were at increased hazard of PM (T-stage and N-stage stratified analysis, log-rank, p<0.0001). Consistent with known clinical outcomes, The Cancer Genome Atlas (TCGA) genomically stable (GS) subtype (log-rank, p=0.0059) and mesenchymal subtype GC were associated with a higher proclivity for peritoneal organotropism (log-rank, p=0.020). Among patients who developed PM, TP53 was the most commonly altered driver mutation (25/51, 49.0%), followed by ARID1A (9/51, 17.6%), CDH1 (8/51, 15.7%), PIK3CA (6/51, 11.8%). Comparing the PTs of patients with PM against those PTs without PM, ELF3, CDH1 and PIGR were significantly associated with PM (all adjusted p<0.05). Amongst TCGA GS tumors, ARID1A mutations (p=0.018) were significantly associated with PM compared to patients with GS tumors with no metastasis. CDH1 mutations were found to significantly co-occur with PIGR and RHOA mutations and was mutually exclusive to TP53. ARID1A mutations were also found to significantly co-occur with multiple other mutations such as PTEN, PIK3CA, ERBB2, and KRAS. Conclusions: CDH1, PIGR, ELF3 mutations in primary GC tumors were found to be significantly associated with metachronous PM.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call