ONO-4578 Plus Nivolumab in Unresectable Advanced or Recurrent Gastric or Gastroesophageal Junction Cancer

Abstract

ONO-4578, an EP4 antagonist, alone and combined with nivolumab, showed acceptable safety profiles and signs of antitumor activity in solid tumors. The expansion part examined the safety, preliminary efficacy, and biomarkers of ONO-4578 plus nivolumab in unresectable advanced or recurrent gastric or gastroesophageal junction (G/GEJ) cancer. Patients were enrolled into three groups: with previous immuno-oncology treatment (IO-treated; n = 30), without IO treatment (IO-naive; n = 30), and with UGT1A1 polymorphism (UGT1A1p; n = 6). Treatment-related adverse events (TRAEs) occurred in 46 patients (grade 3-4 in 17), with no grade 5 events reported. We confirmed the tolerability of the treatment in UGT1A1p. Objective response and disease control rates were 10.0% and 73.3%, respectively, in IO-treated and 16.7% and 40.0%, respectively, in IO-naive. Biomarker analysis indicated immune activation in the tumor microenvironment after the treatment. In conclusion, ONO-4578 plus nivolumab showed a manageable safety profile and antitumor activity in G/GEJ cancer. Trial Registration: Japan Registry of Clinical Trials number: jRCT2080223441; ClinicalTrials.gov identifier: NCT03155061.

Keywords: dose‐limiting toxicity; gastric cancer; nivolumab; phase I; prostaglandin E2 receptor EP4.

FIGURE 1

Therapeutic efficacy of ONO‐4578 plus nivolumab. Maximum percent change in the sum of target lesion diameters from baseline in the IO‐treated (A) and IO‐naive (B) groups. Kaplan–Meier plot of progression‐free survival in the IO‐treated (C) and IO‐naive (D) groups. Kaplan–Meier plot of overall survival in the IO‐treated (E) and IO‐naive (F) groups. The analyses included all patients in each group (n = 30 for each). BOR, best overall response; IO, immuno‐oncology; PD, progressive disease; PR, partial response; SD, stable disease.

FIGURE 2

Pharmacodynamic biomarkers regarding intratumoral T cells in the paired tumor biopsy tissues after ONO‐4578 plus nivolumab treatment. (A) Changes in intratumoral T cell counts from baseline in the paired tumor tissues in the IO‐treated (n = 20) and IO‐naive (n = 10) groups. The number of CD8‐positive cells was evaluated by immunohistochemistry. The values of 0 in the count of CD8‐positive cells (cells/mm²) were replaced with 0.1 in the graph to project the data in a logarithmic scale. (B) Change of T‐effector signature from baseline in the paired tumor tissue in the IO‐treated (n = 20) and IO‐naive (n = 10) groups. The baseline and postdose values were compared using either the Wilcoxon's signed‐rank sum test (A) or the paired t‐test (B). *p < 0.05, ***p < 0.001. BOR, best overall response; IO, immuno‐oncology; PD, progressive disease; PR, partial response; SD, stable disease.

FIGURE 3

Pharmacodynamic biomarkers regarding intratumoral macrophage in the paired tumor biopsy tissues after ONO‐4578 plus nivolumab treatment. (A) Changes in the number of intratumoral CD68‐positive macrophages in the paired tumor tissues were evaluated by immunohistochemistry in the IO‐treated (n = 17) and IO‐naive (n = 9) groups. (B) Changes in the number of intratumoral CD163‐positive macrophages were evaluated by immunohistochemistry in the IO‐treated (n = 20) and IO‐naive (n = 10) groups. (C, D) Change of M1 (C) and M2 (D) macrophage signature scores in the IO‐treated (n = 20) and IO‐naive (n = 10) groups. (E) Changes in the ratio of M1 to M2 macrophages in the paired tumor tissues in the IO‐treated (n = 20) and IO‐naive (n = 10) groups. The M1/M2 macrophage ratio was calculated as the difference in the signature score of M1 macrophages from that of M2 macrophages. The postdose samples were collected on cycle 2 day 15 (1 cycle = 28 days). The baseline and postdose values were compared using either the Wilcoxon's signed‐rank sum test (A, B) or the paired t‐test (C–E). *p < 0.05, **p < 0.01, ***p < 0.001. BOR, best overall response; IO, immuno‐oncology; PD, progressive disease; PR, partial response; SD, stable disease.

FIGURE 4

Biological changes in the paired tumor tissues after ONO‐4578 plus nivolumab treatment. Results for GSEA of the IO‐treated (A) and IO‐naive (B) groups. The Hallmark pathways with a nominal p‐value < 0.05 are shown. The bar color indicates the FDR q‐value. FDR, false discovery rate; GSEA, gene set enrichment analysis; IO, immuno‐oncology.

FIGURE 5

Baseline plasma and urinary PGEM levels were associated with clinical effectiveness. (A) Urinary PGEM levels in the IO‐treated group (PR or SD, n = 22; PD, n = 7). (B) Urinary PGEM levels in the IO‐naive group (PR or SD, n = 12; PD, n = 18). (C) Plasma PGEM levels in IO‐treated group (PR or SD, n = 22; PD, n = 7). (D) Plasma PGEM levels in the IO‐naive group (PR + SD, n = 12; PD, n = 18). The Wilcoxon's rank sum test was used to compare the values in patients with PR or SD and those with PD. (E) Baseline levels of CD8‐positive cells in the tumor tissues of the IO‐treated (n = 25) and IO‐naive groups (n = 15). The number of CD8‐positive cells was evaluated by immunohistochemistry. The Wilcoxon's rank sum test was used to compare the IO‐treated and IO‐naive groups. The values of 0 in the count of CD8‐positive cells (cells/mm²) were replaced with 0.1 in the graph to project the data in a logarithmic scale. *p < 0.05, **p < 0.01, ***p < 0.001. IO, immuno‐oncology; PD, progressive disease; PGEM, prostaglandin E₂ metabolites; PR, partial response; SD, stable disease.