Phase II study of ceralasertib (ATR inhibitor) plus durvalumab in patients with advanced gastric cancer (AGC) who progressed on prior anti-PD-(L)1 therapy.

S Sung Hee Lim (Division of Hematology-Oncology, Department of Medicine, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, South Korea) M Minae An (Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Seoul, South Korea) J Ji Eun Shin M Minsuk Kwon (Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea) S Seung Tae Kim J Jeeyun Lee (Samsung Medical Center, Seoul, South Korea)

Abstract

4062 Background: The standard first-line therapy for advanced gastric cancer (AGC) now includes immune checkpoint inhibitors (ICIs), but resistance is inevitable. ATR inhibition may sensitize tumors to ICIs by enhancing DNA damage and activating the cGAS-STING pathway. We evaluated the efficacy and immunologic impact of ceralasertib (ATRi) plus durvalumab in IO-pretreated AGC. Methods: This open-label phase II trial (NCT03780608) enrolled patients with metastatic gastric adenocarcinoma who progressed on prior anti-PD-(L)1 therapy. Patients received ceralasertib (240mg BID, Days 1–7) and durvalumab (1500mg IV, Day 8) every 28 days. The primary endpoint was ORR (RECIST v1.1). Serial endoscopic biopsies were collected at baseline (BL), post-ATRi monotherapy (FU1), and post-durvalumab combination (FU2) for multi-omics analysis (WES/WTS/scRNA-seq). Results: Thirty patients were enrolled (median 4 prior lines; all HER2-negative). ORR was 33.3% (10/30) and DCR was 56.7%. Median PFS was 4.3 months (95% CI, 1.45–7.10); median OS was not reached. Patients receiving study treatment immediately after prior IO failure had significantly longer PFS than those with intervening non-IO lines (p < 0.05). Grade 3 TRAEs occurred in 10% (3/30) of patients, with no new safety signals identified. Dose reductions of ceralasertib occurred in 4 patients. Multi-omics revealed that higher baseline scarHRD scores correlated with shorter PFS (p = 0.031). scRNA-seq showed that responders (R) had significant depletion of aneuploid clones post-treatment, whereas non-responders (NR) exhibited clonal persistence. Mechanistically, R showed baseline enrichment of the ATR-ATRIP axis, indicating functional dependency on replication stress tolerance. ATR inhibition in R induced replication fork collapse and ATM-CHEK2-mediated apoptosis. Conversely, NR displayed base excision repair deficiency signatures and adaptive G1-phase retention, facilitating evasion of ATRi-induced lethality. Conclusions: Ceralasertib plus durvalumab is efficacious and well-tolerated in IO-refractory AGC. The higher efficacy observed in patients treated immediately after prior IO failure suggests that ATR inhibition may effectively reverse acquired resistance to PD-1 blockade. This biomarker-driven approach warrants further validation in a larger cohort. Clinical trial information: NCT03780608 .

Article Details

Volume / Issue Vol. 44, Issue 16_suppl
Published June 01, 2026
Pages 4062-4062
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (6)

S

Sung Hee Lim

Division of Hematology-Oncology, Department of Medicine, Sungkyunkwan University School of Medicine, Samsung Medical Center, Seoul, South Korea

M

Minae An

Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Seoul, South Korea

J

Ji Eun Shin

M

Minsuk Kwon

Division of Hematology-Oncology, Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea

S

Seung Tae Kim

J

Jeeyun Lee

Samsung Medical Center, Seoul, South Korea