Abstract
A good response to neoadjuvant chemotherapy (NACT) is strongly associated with a higher curative resection rate and favorable outcomes for patients with gastric cancer (GC). We examined the utility of serial circulating tumor DNA (ctDNA) testing for monitoring NACT response and prognosis in stage II-III GC. Seventy-nine patients were enrolled to receive two cycles of NACT following gastrectomy with D2-lymphadenectomy. Plasma at baseline, post-NACT, and after surgery, and tissue at pretreatment and surgery were collected. We used a 425-gene panel to detect genomic alterations (GAs). Results show that the mean cell-free DNA concentration of patients with clinical stage III was significantly higher than patients with stage II (15.43 ng·mL-1 vs 14.40 ng·mL-1 ). After receiving NACT and surgery, the overall detection rate of ctDNA gradually reduced (59.5%, 50.8%, and 47.4% for baseline, post-NACT, and postsurgery). The maximum variant allele frequency (max-VAF) and the number of GAs decreased from 0.50% to 0.08% and from 2.9 to 1.7 after NACT. For patients with a partial response after NACT, the max-VAF and the number of GAs declined significantly, but they increased for patients with progressive disease. Patients with detectable ctDNA at baseline, after NACT, or after surgery have a worse overall survival (OS) than patients with undetectable ctDNA. The estimated 3-year OS was 73% for the post-NACT ctDNA-negative patients and 34% for ctDNA-positive. Patients with perpetual negative ctDNA before and after NACT have the best prognosis. In conclusion, ctDNA was proposed as a potential biomarker to predict prognosis and monitor the NACT response for stage II-III GC patients.