Safety, efficacy, and biomarker analysis from a phase II trial of intensive chemotherapy combined with serplulimab and trastuzumab in patients with advanced HER2-positive gastric cancer.

Abstract

4031 Background: Keynote-811 has proven the efficacy of combined PD-1 and HER2 blockade with chemotherapy in HER2-positive gastric cancer. Our study aims to enhance the survival of patients by replacing standard chemotherapy with intensive chemotherapy and conducting biomarker analysis. Methods: This prospective, single-arm, open-label study was carried out across five centers in China, recruiting patients (pts) with unresectable locally advanced or metastatic HER2-positive gastric cancer. Pts receive Serplulimab (4.5 mg/kg, D1, Q3W), Trastuzumab (initial 8 mg/kg, D1, then 6 mg/kg, D1, Q3W), and the DOS regimen: oxaliplatin (100 mg/m², IV), docetaxel (40 mg/m², IV), and S-1 (40-60mg, BID, D1-14, Q3W). Chemotherapy is up to 8 cycles. Serplulimab and Trastuzumab can be administered until tumor progression. Gastroscopic biopsy was taken before the treatment and dynamic blood samples were collected at C1D1 (T0), C2D1 (T1) and C7D1 (T2) for biomarker analysis. Genomic DNA from tumor tissue and circulating tumor DNA (ctDNA) underwent targeted DNA sequencing containing 571 genes. Results: From July 2022 to September 2024, 40 pts were recruited. The median follow-up was 7.9 months. There were 10 females and 30 males, with a median age of 59 (31-74). 37 pts were eligible for efficacy assessment. The objective response rate (ORR) was 92% (95% CI: 0.87, 0.96), with a complete response rate of 3% (95% CI: 0, 0.05), and a partial response rate of 89% (95% CI: 0.84, 0.94). Median progression free survival has not been reached. 38 pts (95%) experienced adverse events (AEs) of any grade. Grade 3 or above AEs occurred in 14 pts (35%), 2 pts with grade 4 AEs (1 with thrombocytopenia and 1 with myelosuppression). No grade 5 events were observed. The most common AEs were anemia (35%), neutropenia (23%), nausea and vomiting (20%) and leukopenia (20%). 15 (38%) pts had dose interruptions due to AEs, with no treatment discontinuation. HER2 CNV gain was detected in 82.1% (23/28) of tissue samples and 74.3% (26/35) of ctDNA samples. In matched ctDNA and tissue samples (N = 26), 88.5% (23/26) showed concordant HER2 CNV gain results. Elevated tissue HER2 CNV were associated with more pronounced tumor shrinkage (R = -0.35, P = 0.073). Plasma HER2 gain detection rate decreased from 79.1% (19/24) at T0 to 16.7% (4/24) at T1 (P < 0.001). Similar results were observed from T0 to T2 (P < 0.001). Conclusions: This regimen demonstrated remarkable efficacy with a high ORR and manageable toxicity. Tumor HER2 CNV and ctDNA dynamic monitoring correlated with treatment efficacy. The subsequent results of biomarker analysis will be presented at the upcoming conference. Clinical trial information: NCT05311189 . Pts characteristics. N=40 Age, ≥65 years 38% Male 75% PD-L1 status  CPS ≥1 65%  CPS <1 13%  Unknown 22% HER2 status  IHC 2+ ISH positive 25%  IHC 3+ 75% Primary gastrectomy or esophagectomy  Yes 15%  No 85% Metastatic sites  0–2 60%  ≥3 40%