Longitudinal circulating tumor DNA (ctDNA) analysis during treatment (Tx) of locally advanced resectable (LAR) gastric or gastroesophageal junction (G/GEJ) adenocarcinoma (ADENOCA): The PLAGAST prospective biomarker study.

4028 Background: After surgery, patients (pts) with LAR G/GEJ ADENOCA have a high recurrence risk despite Tx. Here, we investigate the ability of ctDNA to predict Tx response and improve risk stratification in a sub-cohort of pts from the PLAGAST prospective study (NCT02674373). Methods: Plasma samples were prospectively collected at baseline (before neoadjuvant therapy (NAT)/before surgery for NAT-naive pts), during-NAT, post-NAT, and post-surgery within 2-12 weeks, before adjuvant Tx molecular residual disease (MRD) window. Retrospective ctDNA analysis was performed using a clinically validated personalized, tumor-informed 16-plex PCR assay (Signatera, Natera Inc.). ctDNA results were evaluated for their correlation to recurrence-free survival (RFS) and overall survival (OS). Results: Between 07/2015 and 12/2022,plasma samples (n=283) were collected from 63 pts with LAR (≥cT2 and/or cN+), G (N=38/63; 60%) and GEJ (N=25/63; 40%) ADENOCA, median (m) age: 66 years (range: 34-86), and mfollow-up: 31 months (mo) (range: 2-93). ctDNA-clearance from baseline to during-NAT (N=30) or post-NAT (N=28), and MRD (N=50) window was associated with longer RFS and OS. Persistent ctDNA positivity (+ve) during-NAT was associated with a mRFS of 13.3mo vs. not reached (NR) for pts who achieved ctDNA-clearance (HR=4; p=0.032), and a mOS of NR for both groups, with 24mo OS rate at 95.6% vs. 58.6% (HR=5.85; p=0.097), respectively. ctDNA clearance post-NAT was associated with a mRFS of 7.79 vs. 52.04 mo (HR= 3.69; p=0.019) and a mOS of 18.9mo vs. NR (HR=8.78; p=0.007) for ctDNA-cleared vs. pts who did not clear, respectively. During the MRD window, mRFS of 3.57mo vs. NR (HR=12.94; p<0.0001) and mOS of 8.59mo vs. NR (HR=14.54; p<0.0001) was observed for ctDNA+ve and ctDNA negative pts, respectively. ctDNA dynamics revealed pts with early ctDNA clearance (during NAT) had the best outcomes, followed by pts who cleared ctDNA post-NAT, while pts who remained ctDNA+ve at baseline, during-NAT, and post-NAT had worse RFS and OS (24mo RFS rate of 75% vs. 48% vs. NR and 24mo OS rate of 100% vs. 80% vs. 17%, respectively). Of note, pts who achieved ctDNA clearance post-NAT correlated with tumor regression grade (TRG) 1/2/3 when compared to a TRG score of 4/5 in pts who were unable to clear ctDNA post-NAT (p=0.0345). Conclusions: Tumor-informed ctDNA results were predictive of NAT response and correlated with long-term outcomes with early ctDNA clearance associated with superior RFS and OS; meanwhile persistent ctDNA+ve post-NAT/pre-surgery/during MRD correlated with inferior OS and/or RFS as well as poor pathological response. In this initial analysis of the PLAGAST study, longitudinal ctDNA monitoring was prognostic of patient outcomes and may guide NAT clinical decision-making in pts with LAR G/GEJ ADENOCA. Clinical trial information: NCT02674373 .