ABSTRACT
Background
The CheckMate 649 trial showed the superiority of first‐line nivolumab plus chemotherapy (Nivo‐CT) over chemotherapy alone for advanced gastric or gastroesophageal junction (GEJ) cancer with a programmed cell death‐ligand 1 (PD‐L1) combined positive score (CPS) ≥ 5. However, few studies have investigated the correlation between PD‐L1 CPS and the efficacy of first‐line Nivo‐CT in the real‐world setting.
Methods
This multicenter retrospective study included patients with unresectable advanced or recurrent gastric or GEJ cancer who were started on first‐line Nivo‐CT between December 2021 and December 2023 at any of four institutions in Japan. PD‐L1 CPS was classified into three groups (< 1, 1 to < 5, and ≥ 5) before the start of Nivo‐CT.
Results
A total of 234 patients were included (median age 70 years; 64.5% male; 89.7% with gastric primary). PD‐L1 CPS was < 1 in 35 patients (15.0%), 1 to < 5 in 76 (32.5%), and ≥ 5 in 123 (52.5%). Median follow‐up was 14.0 months (range 2.3–30.3). Median overall survival was 18.7 (95% confidence interval [CI] 17.4–not reached), 23.4 (95% CI 17.7–not reached), and 18.3 months (95% CI 13.3–22.5), with objective response rates of 30.8%, 66.7%, and 61.3%. Multivariate analysis did not identify PD‐L1 CPS as a prognostic factor, whereas an Eastern Cooperative Oncology Group performance status ≥ 1, a GEJ primary, and two or more metastatic sites were associated with poor prognosis.
Conclusions
The PD‐L1 CPS had no significant effect on prognosis in patients with unresectable advanced or recurrent gastric or GEJ cancer treated with first‐line Nivo‐CT.
Keywords: combined positive score, gastric cancer, nivolumab plus chemotherapy
1. Introduction
Gastric cancer is the fifth most commonly diagnosed cancer and the fourth leading cause of cancer‐related mortality worldwide. In several Asian countries, gastric cancer is the most common cancer and the leading cause of cancer‐related deaths in men [1]. Fluoropyrimidine plus platinum‐based chemotherapy has long been the first‐line treatment for unresectable advanced or recurrent human epidermal growth factor receptor 2 (HER2)‐negative gastric or gastroesophageal junction (GEJ) cancer, but results in poor median overall survival (OS) of around 12 months [2, 3, 4, 5].
Nivolumab is a human monoclonal anti‐programmed death‐1 (PD‐1) antibody that selectively blocks the interaction of PD‐1 on activated T‐cells and its ligands with immune and tumor cells, thereby preventing immunosuppression and resulting in immunogenic cell death [6]. The combined positive score (CPS), determined by expression of programmed death‐ligand 1 (PD‐L1) on tumor cells and tumor‐associated immune cells, is a biomarker that predicts the efficacy of immune checkpoint inhibitors (ICIs) in patients with advanced gastric or GEJ cancer [7, 8, 9]. The phase III CheckMate 649 trial demonstrated that nivolumab plus chemotherapy (Nivo‐CT) was superior to chemotherapy alone in terms of OS and progression‐free survival (PFS) as a first‐line treatment for advanced HER2‐negative gastric or GEJ cancer in patients with a PD‐L1 CPS of ≥ 5 [10]. Nivo‐CT was also superior to chemotherapy alone in patients with a PD‐L1 CPS of ≥ 1 and in the overall population but not in those with a PD‐L1 CPS of < 1 or < 5. Consistent with the CheckMate 649 trial, subgroup analyses of the PD‐L1 CPS in the KEYNOTE‐859 and RATIONALE‐305 trials demonstrated survival benefits from anti‐PD‐1 antibodies as the PD‐L1 CPS increased [11, 12].
Based on the results of CheckMate 649, the American Society of Clinical Oncology guideline and the European Society For Medical Oncology living guideline recommend Nivo‐CT as a first‐line treatment for advanced HER2‐negative gastric cancer if the PD‐L1 CPS is ≥ 5, on a case‐by‐case basis for a PD‐L1 CPS of 1–4, and chemotherapy alone if the PD‐L1 CPS is < 1 [13, 14]. The National Comprehensive Cancer Network guideline lists Nivo‐CT as a preferred regimen for patients with a PD‐L1 CPS ≥ 1, whereas the Pan‐Asian‐adapted European Society For Medical Oncology clinical practice guideline recommends Nivo‐CT only when the PD‐L1 CPS is ≥ 5 [15, 16]. These guidelines recommend first‐line Nivo‐CT for a PD‐L1 CPS ≥ 5, but its use when the PD‐L1 CPS is 1 to < 5 remains controversial. Patients with PD‐L1 CPS 1 to < 5 represent a clinically important subgroup because treatment decisions remain challenging owing to variable guideline recommendations and the modest benefit observed in this range.
Microsatellite instability (MSI), a hypermutable phenotype arising from defects in the DNA mismatch repair (MMR) pathway, has emerged as a robust predictive biomarker of the response to ICIs across multiple types of cancer [17, 18]. MSI‐high (MSI‐H) or deficient MMR (dMMR) tumors have been reported to be highly responsive to ICIs, and both CheckMate 649 and KEYNOTE‐859 demonstrated that ICI‐containing treatments yield consistently favorable outcomes in this patient population [10, 11].
These findings indicate that the PD‐L1 CPS and MSI status are reliable biomarkers for prediction of the survival benefit of adding nivolumab to chemotherapy in first‐line treatment. However, real‐world evidence for the ability of the PD‐L1 CPS and MSI status to predict the effectiveness of first‐line Nivo‐CT in patients with HER2‐negative unresectable advanced or recurrent gastric or GEJ cancer remains scarce. Therefore, in this study, we retrospectively investigated the correlations of the PD‐L1 CPS and MSI status with the clinical outcomes of first‐line Nivo‐CT in this population.
2. Methods
2.1. Patients
This multicenter retrospective observational study included patients with gastric or GEJ cancer who received Nivo‐CT as a first‐line treatment at any of four participating Japanese institutions. The inclusion criteria were as follows: age ≥ 18 years; histologically confirmed HER2‐negative unresectable advanced or recurrent gastric/GEJ adenocarcinoma; no previous systemic therapy for unresectable or metastatic disease (previous adjuvant or neoadjuvant chemotherapy was permitted if disease progression occurred ≥ 6 months after completion of treatment); initiation of Nivo‐CT between December 2021 and December 2023; and results of PD‐L1 CPS assessment available. Patients who had previously received immunotherapy with anti‐PD‐1, anti‐PD‐L1, or anti‐CTLA‐4 antibodies were excluded.
We reviewed medical records and collected baseline characteristics of patients prior to the initiation of Nivo‐CT. PD‐L1 CPS was classified into three groups (< 1, 1 to < 5, and ≥ 5) at each center using the PD‐L1 immunohistochemistry 28‐8 pharmDx assay (Dako; Agilent Technologies, Santa Clara, CA). The study was approved by the ethics committees of all participating institutions (National Cancer Center Hospital, No. 2023‐266; Kanagawa Cancer Center, No. 2024‐29; Saitama Cancer Center, No. 1831; Ishikawa Prefectural Central Hospital, No. 118). The need for written informed consent was waived. However, information about the study was posted on the institutional websites to provide patients with the opportunity to opt out.
2.2. Treatment and Assessment
Patients received Nivo‐CT, consisting of Nivo‐SOX (nivolumab 360 mg on Day 1, S‐1 [tegafur–gimeracil–oteracil potassium] 40 mg/m2 twice daily on Days 1–14, and oxaliplatin 130 mg/m2 on Day 1, every 3 weeks), Nivo‐CAPOX (nivolumab 360 mg on Day 1, capecitabine 1000 mg/m2 twice daily on Days 1–14, and oxaliplatin 130 mg/m2 on Day 1, every 3 weeks), or Nivo‐FOLFOX (nivolumab 240 mg on Day 1, 5‐fluorouracil 400 mg/m2 on Day 1 and 1200 mg/m2 on Days 1 and 2, l‐leucovorin 200 mg/m2 on Day 1, and oxaliplatin 85 mg/m2 on Day 1, every 2 weeks). Patients received the treatment until disease progression, clinical deterioration, unacceptable toxicity, or discontinuation at the patient's request. Dose modifications and temporary suspensions were permitted.
The neutrophil‐to‐lymphocyte ratio (NLR) was calculated by dividing the neutrophil count by the lymphocyte count. Efficacy was evaluated by OS, PFS, and the objective response rate (ORR). OS was defined as the interval between initiation of Nivo‐CT and death from any cause or censored at the last date of confirmed survival. PFS was defined as the interval between initiation of Nivo‐CT and disease progression or death from any cause, whichever occurred first; survivors without disease progression were censored on the last date of follow‐up. The data cut‐off date was June 30, 2024. ORR was evaluated according to the RECIST version 1.1 in patients who had target lesions. The Common Terminology Criteria for Adverse Events (Ver. 5.0) was used to evaluate adverse events (AEs). For AEs other than immune‐related AEs (irAEs), only those that were grade ≥ 3 were collected.
2.3. Statistical Analysis
Categorical variables were compared between groups using Pearson's chi‐squared test or Fisher's exact test, and continuous variables were compared using the Kruskal–Wallis test. OS and PFS were estimated using the Kaplan–Meier method and compared using the log‐rank test. Univariable analysis and multivariable analysis using Cox regression models were performed to evaluate the association between the PD‐L1 CPS and survival outcomes. All statistical analyses were conducted using R software (version 4.4.1; R Foundation for Statistical Computing, Vienna, Austria). All tests were two‐sided, and a p value of < 0.05 was considered to be statistically significant.
3. Results
3.1. Patient Characteristics
A total of 234 patients were included. The median age was 70.5 years (range 18–84) and 151 patients (64.5%) were men. Eastern Cooperative Oncology Group (ECOG) performance status (PS) was 0 in 114 patients (51.3%). PD‐L1 CPS was < 1 in 35 patients (15.0%), 1 to < 5 in 76 (32.5%), and ≥ 5 in 123 (52.5%). PD‐L1 CPS was assessed using primary tumor specimens in most cases (226/234, 96.6%). Nivo‐SOX was selected as the Nivo‐CT regimen for 153 patients (65.4%), Nivo‐CAPOX for 11 (4.7%), and Nivo‐FOLFOX for 70 (29.9%). Patient characteristics are shown according to PD‐L1 CPS status in Table 1. Liver metastases were found in 43 (35.0%) of the patients with a PD‐L1 CPS ≥ 5, which was a significantly higher rate than that in patients with a PD‐L1 CPS of < 1 or 1 to < 5 (p = 0.001). With the exception of liver metastasis, there were no significant differences in the clinical profiles of the three groups.
TABLE 1.
Patient characteristics according to PD‐L1 CPS status.
| Total, N = 234 | PD‐L1 CPS < 1 n = 35 | 1 ≤ PD‐L1 CPS < 5 n = 76 | PD‐L1 CPS ≥ 5 n = 123 | p |
|---|---|---|---|---|
| Median age (range) (years) | 68 (34–83) | 70 (30–84) | 72 (18–84) | 0.253 |
| Age (years) | 0.478 | |||
| < 65 | 14 (40.0%) | 25 (32.9%) | 36 (29.3%) | |
| ≥ 65 | 21 (60.0%) | 51 (67.1%) | 87 (70.7%) | |
| Sex | 0.386 | |||
| Male | 19 (54.3%) | 50 (65.8%) | 82 (66.7%) | |
| Female | 16 (45.7%) | 26 (34.2%) | 41 (33.3%) | |
| ECOG PS | 0.451 | |||
| 0 | 15 (42.9%) | 38 (50.0%) | 67 (54.5%) | |
| ≥ 1 | 20 (57.1%) | 38 (50.0%) | 56 (45.5%) | |
| Primary tumor | 0.730 | |||
| Gastric | 33 (94.3%) | 68 (89.5%) | 109 (88.6%) | |
| GEJ | 2 (5.7%) | 8 (10.5%) | 14 (11.4%) | |
| Status | 0.920 | |||
| Unresectable advanced | 28 (80.0%) | 60 (78.9%) | 100 (81.3%) | |
| Recurrent | 7 (20.0%) | 16 (21.1%) | 23 (18.7%) | |
| Metastatic sites, n | 0.175 | |||
| 0–1 | 22 (62.9%) | 55 (72.4%) | 73 (59.3%) | |
| ≥ 2 | 13 (37.1%) | 21 (27.6%) | 50 (40.7%) | |
| Liver metastasis | 0.001 | |||
| No | 30 (85.7%) | 65 (85.5%) | 80 (65.0%) | |
| Yes | 5 (14.3%) | 11 (14.5%) | 43 (35.0%) | |
| Lung metastasis | 0.555 | |||
| No | 32 (91.4%) | 71 (93.4%) | 117 (95.1%) | |
| Yes | 3 (8.6%) | 5 (6.6%) | 6 (4.9%) | |
| Lymph node metastasis | 0.462 | |||
| No | 20 (57.1%) | 38 (50.0%) | 56 (45.5%) | |
| Yes | 15 (42.9%) | 38 (50.0%) | 67 (54.5%) | |
| Peritoneal metastasis | 0.720 | |||
| No | 17 (48.6%) | 43 (56.6%) | 68 (55.3%) | |
| Yes | 18 (51.4%) | 33 (43.4%) | 55 (44.7%) | |
| Ascites | 0.914 | |||
| No | 23 (65.7%) | 53 (69.7%) | 84 (68.3%) | |
| Yes | 12 (34.3%) | 23 (30.3%) | 39 (31.7%) | |
| Histology | 0.740 | |||
| Intestinal | 10 (28.6%) | 27 (35.5%) | 38 (30.9%) | |
| Diffuse | 25 (71.4%) | 49 (64.5%) | 82 (66.7%) | |
| Unknown | 0 (0.0%) | 0 (0.0%) | 3 (2.4%) | |
| Signet‐ring cell carcinoma | 0.812 | |||
| No | 28 (80.0%) | 64 (84.2%) | 104 (84.6%) | |
| Yes | 7 (20.0%) | 12 (15.8%) | 19 (15.4%) | |
| ALP | 0.265 | |||
| Normal | 27 (77.1%) | 48 (63.2%) | 88 (71.5%) | |
| High | 8 (22.9%) | 28 (36.8%) | 35 (28.5%) | |
| MSI status | 0.178 | |||
| MSI‐H/dMMR | 0 (0.0%) | 1 (1.3%) | 5 (4.1%) | |
| MSS/pMMR | 17 (48.6%) | 36 (47.4%) | 37 (30.1%) | |
| Unknown | 18 (51.4%) | 39 (51.3%) | 81 (65.8%) | |
| Chemotherapy regimen | 0.771 | |||
| SOX | 20 (57.1%) | 53 (69.7%) | 80 (65.0%) | |
| CAPOX | 2 (5.7%) | 3 (3.9%) | 6 (4.9%) | |
| FOLFOX | 13 (37.1%) | 20 (26.3%) | 37 (30.1%) | |
| Second line treatment | 0.701 | |||
| No | 18 (51.4%) | 45 (59.2%) | 67 (54.5%) | |
| Yes | 17 (48.6%) | 31 (40.8%) | 56 (45.5%) |
Abbreviations: ALP, alkaline phosphatase; CAPOX, capecitabine and oxaliplatin; dMMR, deficient mismatch repair; ECOG PS, Eastern Cooperative Oncology Group performance status; FOLFOX, 5‐fluorouracil, l‐leucovorin, and oxaliplatin; GEJ, gastroesophageal junction; MSI‐H, microsatellite instability‐high; MSS, microsatellite stable; PD‐L1 CPS, programmed death‐ligand 1 combined positive score; pMMR, proficient mismatch repair; SOX, tegafur–gimeracil–oteracil potassium and oxaliplatin.
3.2. Clinical Outcomes in All Patients
At the time of data cut‐off, the median follow‐up duration was 14.0 months (range 2.3–30.3). The median OS was 18.7 months (95% confidence interval [CI] 17.5–23.4) and the median PFS was 7.7 months (95% CI 6.3–9.7). Among the 101 patients (43.2%) with target lesions, a complete response was achieved in 6 (5.9%) and a partial response in 61 (60.4%), resulting in an ORR of 66.3% (95% CI 56.7–74.8). Of the 234 patients included in the study, 55 (23.5%) were still receiving first‐line Nivo‐CT at the data cut‐off, and 126 (53.8%) received at least one subsequent therapy for unresectable advanced or recurrent gastric or GEJ cancer. The details of subsequent therapies according to PD‐L1 CPS are provided in Table S1.
3.3. Effect of PD‐L1 CPS on Clinical Outcomes
The median OS was 18.7 (95% CI 17.4–not reached [NR]), 23.4 (95% CI 17.7–NR), and 18.3 months (95% CI 13.3–22.5) in patients with a PD‐L1 CPS of < 1, 1 to < 5, and ≥ 5, respectively, with a respective median PFS of 9.7 (95% CI 5.9–16.3), 8.8 (95% CI 7.2–13.0), and 6.3 months (95% CI 5.3–9.0) (Figure 1) and respective ORRs of 50.0% (95% CI 21.5–78.5), 79.1% (95% CI 59.1–91.2), and 63.8% (95% CI 52.0–74.1). Figure 2A,B shows the ORR and waterfall plot indicating the best responses to Nivo‐CT.
FIGURE 1.
Effectiveness outcomes. (A) OS and (B) PFS. CI, confidence interval; OS, overall survival; PD‐L1 CPS, programmed death‐ligand 1 combined positive score; PFS, progression‐free survival.
FIGURE 2.
(A) Objective response rate and (B) waterfall plot indicating best responses to nivolumab plus chemotherapy. CR, complete response; NE, not evaluable; PD, progressive disease; PD‐L1 CPS, programmed death‐ligand 1 combined positive score; PR, partial response; SD, stable disease.
In the analysis stratified by PD‐L1 CPS status (≥ 5 vs. < 5), there was no significant difference in median OS between the two groups (18.3 vs. 18.8 months; hazard ratio [HR] 1.36, 95% CI 0.91–2.02, p = 0.129). Similarly, there was no significant difference in median PFS between these groups (6.3 vs. 9.3 months; HR 1.25, 95% CI 0.92–1.70, p = 0.149). Among patients without liver metastasis, median OS was also comparable between patients with PD‐L1 CPS ≥ 5 and those with PD‐L1 CPS < 5 (20.3 vs. 18.8 months; HR 1.14, 95% CI 0.70–1.85, p = 0.591). Likewise, no significant difference in median PFS was observed between the two groups (7.6 vs. 9.7 months; HR 1.13, 95% CI 0.78–1.64, p = 0.529). The ORR was 63.8% (95% CI 52.0–74.1) in patients with a PD‐L1 CPS of ≥ 5 and 71.9% (95% CI 54.5–84.6) in those with a PD‐L1 CPS of < 5 (p = 0.422). The patients were then stratified according to whether the CPS was ≥ 1 or < 1 to evaluate further the prognostic significance of PD‐L1 expression. There was no significant difference in median OS between the group with a CPS of ≥ 1 and the group with a CPS of < 1 (18.8 vs. 18.7 months; HR 1.13, 95% CI 0.65–1.97, p = 0.651) or in median PFS (7.3 vs. 9.7 months; HR 1.09, 95% CI 0.71–1.67, p = 0.699). The ORR was 67.7% in the patients with a PD‐L1 CPS of ≥ 1 (95% CI 57.7–76.4) and 50.0% in those with a PD‐L1 CPS of < 1 (95% CI 21.5–78.5) (p = 0.308).
In univariate analysis with age (≥ 65 or < 65 years), ECOG PS (≥ 1 or 0), primary site (GEJ or gastric cancer), status (recurrent or unresectable advanced), number of metastatic sites (≥ 2 or < 2), liver metastasis (yes or no), NLR (≥ 3 or < 3), serum alkaline phosphatase level (high or normal), and PD‐L1 CPS status (≥ 1 or < 1) as covariates, an ECOG PS ≥ 1, a GEJ primary, ≥ 2 metastatic sites, liver metastasis, NLR ≥ 3, and a high alkaline phosphatase level were significantly associated with shorter OS and PFS. Multivariate analysis demonstrated that ECOG PS (HR 1.96, 95% CI 1.30–2.95, p = 0.001), a GEJ primary (HR 2.09, 95% CI 1.15–3.78, p = 0.015), and ≥ 2 metastatic sites (HR 1.63, 95% CI 1.06–2.51, p = 0.025) were associated with poor OS, while ECOG PS (HR 1.79, 95% CI 1.31–2.46, p < 0.001) and a GEJ primary (HR 2.10, 95% CI 1.31–3.38, p = 0.002) were associated with poor PFS; PD‐L1 CPS was not a prognostic factor in either OS or PFS (Tables 2A and 2B).
TABLE 2A.
Univariable and multivariable analyses for overall survival with Cox regression models.
| Variable | Univariable analysis | Multivariable analysis | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | p | HR | 95% CI | p | |
| Age (years) | ||||||
| < 65 | — | — | — | — | ||
| ≥ 65 | 0.989 | 0.655–1.492 | 0.956 | 1.060 | 0.694–1.617 | 0.789 |
| ECOG PS | ||||||
| 0 | — | — | — | — | ||
| ≥ 1 | 2.066 | 1.384–3.082 | < 0.001 | 1.962 | 1.303–2.953 | 0.001 |
| Primary tumor | ||||||
| Gastric | — | — | — | — | ||
| GEJ | 1.896 | 1.072–3.355 | 0.028 | 2.085 | 1.152–3.776 | 0.015 |
| Status | ||||||
| Unresectable advanced | — | — | — | — | ||
| Recurrent | 0.774 | 0.459–1.307 | 0.338 | 0.873 | 0.506–1.508 | 0.627 |
| Metastatic sites, n | ||||||
| 0–1 | — | — | — | — | ||
| ≥ 2 | 1.940 | 1.306–2.880 | 0.001 | 1.633 | 1.063–2.508 | 0.025 |
| Liver metastasis | ||||||
| No | — | — | — | — | ||
| Yes | 1.573 | 1.036–2.387 | 0.033 | 1.095 | 0.687–1.745 | 0.702 |
| NLR | ||||||
| < 3 | — | — | — | — | ||
| ≥ 3 | 1.936 | 1.289–2.907 | 0.001 | 1.500 | 0.978–2.300 | 0.063 |
| ALP | ||||||
| Normal | — | — | — | — | ||
| High | 1.982 | 1.323–2.970 | < 0.001 | 1.506 | 0.975–2.326 | 0.065 |
| PD‐L1 CPS | ||||||
| < 1 | — | — | — | — | ||
| ≥ 1 | 1.134 | 0.654–1.965 | 0.654 | 1.059 | 0.609–1.841 | 0.840 |
Abbreviations: ALP, alkaline phosphatase; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; GEJ, gastroesophageal junction; HR, hazard ratio; NLR, neutrophil‐to‐lymphocyte ratio; PD‐L1 CPS, programmed death‐ligand 1 combined positive score.
TABLE 2B.
Univariable and multivariable analyses of progression‐free survival with Cox regression models.
| Variable | Univariable analysis | Multivariable analysis | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | p | HR | 95% CI | p | |
| Age (years) | ||||||
| < 65 | — | — | — | — | ||
| ≥ 65 | 1.100 | 0.794–1.522 | 0.567 | 1.076 | 0.770–1.505 | 0.668 |
| ECOG PS | ||||||
| 0 | — | — | — | — | ||
| ≥ 1 | 1.740 | 1.280–2.367 | < 0.001 | 1.791 | 1.305–2.457 | < 0.001 |
| Primary tumor site | ||||||
| Gastric | — | — | — | — | ||
| GEJ | 1.912 | 1.203–3.039 | 0.006 | 2.103 | 1.307–3.383 | 0.002 |
| Status | ||||||
| Unresectable advanced | — | — | — | — | ||
| Recurrent | 1.118 | 0.766–1.631 | 0.564 | 1.201 | 0.812–1.775 | 0.359 |
| Metastatic sites, n | ||||||
| 0–1 | — | — | — | — | ||
| ≥ 2 | 1.551 | 1.137, 2.116 | 0.006 | 1.253 | 0.890, 1.764 | 0.196 |
| Liver metastasis | ||||||
| No | — | — | — | — | ||
| Yes | 1.663 | 1.189–2.327 | 0.003 | 1.342 | 0.925–1.949 | 0.122 |
| NLR | ||||||
| < 3 | — | — | — | — | ||
| ≥ 3 | 1.396 | 1.027–1.898 | 0.033 | 1.159 | 0.838–1.602 | 0.373 |
| ALP | ||||||
| Normal | — | — | — | — | ||
| High | 1.678 | 1.213–2.320 | 0.002 | 1.376 | 0.973–1.945 | 0.071 |
| PD‐L1 CPS | ||||||
| < 1 | — | — | — | — | ||
| ≥ 1 | 1.088 | 0.710–1.665 | 0.699 | 1.007 | 0.654–1.549 | 0.976 |
Abbreviations: ALP, alkaline phosphatase; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; GEJ, gastroesophageal junction; HR, hazard ratio; NLR, neutrophil‐to‐lymphocyte ratio; PD‐L1 CPS, programmed death‐ligand 1 combined positive score.
3.4. Clinical Outcomes in Patients With MSI‐H or dMMR Tumors
Six patients (2.6%) were identified as having MSI‐H or dMMR tumors. Among these, five patients had a PD‐L1 CPS of ≥ 5 and one had a PD‐L1 CPS of 1 to < 5; no patient had a PD‐L1 CPS of < 1. The median OS and PFS in patients with MSI‐H or dMMR were not reached at the time of analysis (OS 95% CI NR–NR; PFS 95% CI 6.7–NR) (Figure S1A,B). One‐year OS and PFS rates were 80.0% (95% CI 51.6–100.0) and 62.5% (95% CI 32.0–100.0), respectively. Among the four patients with measurable lesions, three achieved a partial response and one achieved stable disease, yielding an ORR of 75.0%. In the analysis stratified by MSI status (MSI‐H/dMMR or MSS/pMMR), there was no significant difference in median OS (NR vs. 18.7 months; HR 0.47, 95% CI 0.06–3.42, p = 0.443) or in median PFS (NR vs. 7.7 months; HR 0.40, 95% CI 0.10–1.62, p = 0.182).
3.5. Toxicity
In the overall population, grade ≥ 3 AEs other than irAEs were observed in 74 patients (31.6%) and any grade irAEs in 70 (29.9%) (Table 3). The most common grade ≥ 3 AEs other than irAEs were neutropenia (n = 37, 15.8%), decreased appetite (n = 10, 4.3%), and peripheral sensory neuropathy (n = 9, 3.8%). The most common irAEs were pneumonia (n = 16, 6.8%), hypothyroidism (n = 13, 5.6%), pruritus (n = 8, 3.4%), and rash (n = 8, 3.4%). Grade ≥ 3 irAEs included pneumonia (n = 5, 2.1%), hypopituitarism (n = 5, 2.1%), increased aspartate/alanine aminotransaminase levels (n = 3, 1.3%), and colitis (n = 3, 1.3%). There were no treatment‐related deaths.
TABLE 3.
Adverse events.
| Overall N = 234 | PD‐L1 CPS < 1 n = 35 | 1 ≤ PD‐L1 CPS < 5 n = 76 | PD‐L1 CPS ≥ 5 n = 123 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Grade 1–2 | Grade 3 | Grade 4 | Grade 1–2 | Grade 3 | Grade 4 | Grade 1–2 | Grade 3 | Grade 4 | Grade 1–2 | Grade 3 | Grade 4 | |
| AEs except for irAEs | ||||||||||||
| Neutropenia | — | 32 (13.7%) | 5 (2.1%) | — | 2 (5.7%) | 1 (2.9%) | — | 13 (17.1%) | 2 (2.6%) | — | 17 (13.8%) | 2 (1.6%) |
| Decreased appetite | — | 10 (4.3%) | 0 (0.0%) | — | 0 (0.0%) | 0 (0.0%) | — | 3 (3.9%) | 0 (0.0%) | — | 7 (5.7%) | 0 (0.0%) |
| Peripheral sensory neuropathy | — | 9 (3.8%) | 0 (0.0%) | — | 1 (2.9%) | 0 (0.0%) | — | 2 (2.6%) | 0 (0.0%) | — | 6 (4.9%) | 0 (0.0%) |
| Thrombocytopenia | — | 7 (3.0%) | 0 (0.0%) | — | 2 (5.7%) | 0 (0.0%) | — | 2 (2.6%) | 0 (0.0%) | — | 3 (2.4%) | 0 (0.0%) |
| Anemia | — | 5 (2.1%) | 2 (0.9%) | — | 0 (0.0%) | 0 (0.0%) | — | 4 (5.3%) | 0 (0.0%) | — | 1 (0.8%) | 2 (1.6%) |
| Leukopenia | — | 5 (2.1%) | 0 (0.0%) | — | 0 (0.0%) | 0 (0.0%) | — | 2 (2.6%) | 0 (0.0%) | — | 3 (2.4%) | 0 (0.0%) |
| irAEs | ||||||||||||
| Pneumonia | 11 (4.7%) | 4 (1.7%) | 1 (0.4%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 2 (2.6%) | 3 (3.9%) | 1 (1.3%) | 9 (7.3%) | 1 (0.8%) | 0 (0.0%) |
| Hypothyroidism | 13 (5.6%) | 0 (0.0%) | 0 (0.0%) | 1 (2.9%) | 0 (0.0%) | 0 (0.0%) | 3 (3.9%) | 0 (0.0%) | 0 (0.0%) | 9 (7.3%) | 0 (0.0%) | 0 (0.0%) |
| Hypopituitarism | 3 (1.3%) | 5 (2.1%) | 0 (0.0%) | 0 (0.0%) | 2 (5.7%) | 0 (0.0%) | 2 (2.6%) | 2 (2.6%) | 0 (0.0%) | 1 (0.8%) | 1 (0.8%) | 0 (0.0%) |
| Pruritus | 8 (3.4%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 1 (1.3%) | 0 (0.0%) | 0 (0.0%) | 7 (5.7%) | 0 (0.0%) | 0 (0.0%) |
| Rash | 8 (3.4%) | 0 (0.0%) | 0 (0.0%) | 2 (5.7%) | 0 (0.0%) | 0 (0.0%) | 3 (3.9%) | 0 (0.0%) | 0 (0.0%) | 3 (2.4%) | 0 (0.0%) | 0 (0.0%) |
| Increased AST/ALT | 4 (1.7%) | 0 (0.0%) | 3 (1.3%) | 1 (2.9%) | 0 (0.0%) | 0 (0.0%) | 1 (1.3%) | 0 (0.0%) | 2 (2.6%) | 2 (1.6%) | 0 (0.0%) | 1 (0.8%) |
| Adrenal insufficiency | 6 (2.6%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 1 (1.3%) | 0 (0.0%) | 0 (0.0%) | 5 (4.1%) | 0 (0.0%) | 0 (0.0%) |
| Colitis | 2 (0.9%) | 3 (1.3%) | 0 (0.0%) | 1 (2.9%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 1 (1.3%) | 0 (0.0%) | 1 (0.8%) | 2 (1.6%) | 0 (0.0%) |
| Hyperthyroidism | 4 (1.7%) | 0 (0.0%) | 0 (0.0%) | 1 (2.9%) | 0 (0.0%) | 0 (0.0%) | 2 (2.6%) | 0 (0.0%) | 0 (0.0%) | 1 (0.8%) | 0 (0.0%) | 0 (0.0%) |
Abbreviations: AEs, adverse events; ALT, alanine transaminase; AST, aspartate transaminase; irAEs, immune‐related adverse events; PD‐L1 CPS, programmed death‐ligand 1 combined positive score.
4. Discussion
This large‐scale, multicenter retrospective study investigated the outcomes of first‐line Nivo‐CT in 234 Japanese patients with HER2‐negative unresectable advanced or recurrent gastric or GEJ cancer according to their PD‐L1 CPS. We found that a high PD‐L1 CPS was not associated with improved OS or PFS. Moreover, multivariate analysis identified an ECOG PS ≥ 1, a GEJ primary, and ≥ 2 metastatic sites to be independent poor prognostic factors. In patients with MSI‐H/dMMR tumors, median OS and PFS were not reached at the time of analysis, and the ORR was 75.0%.
The median OS of 18.7 months, PFS of 7.7 months, and ORR of 66.3% in our overall study population are comparable with the results of the CheckMate 649 trial (13.8 months, 7.7 months, and 58%, respectively), with OS being relatively favorable [10]. The ATTRACTION‐4 trial, a phase III study that evaluated the efficacy of first‐line Nivo‐CT in Asian patients, reported an OS of 17.5 months, a PFS of 10.5 months, and an ORR of 57.5% in the overall population [19]. It has been suggested that the high frequency of subsequent anticancer therapy (71.5%) partly contributed to the favorable prognosis of patients in ATTRACTION‐4. In our study, 23.5% of patients remained on first‐line Nivo‐CT, and 53.8% had received at least one subsequent anticancer therapy by the date of data cutoff; these rates are higher than those in CheckMate 649 (10.7% and 37.6%, respectively) and may have contributed to our favorable survival outcomes. Furthermore, the efficacy of first‐line Nivo‐CT and the higher rates of subsequent chemotherapy in our study were those in other real‐world studies in Japan [20, 21].
In the CheckMate 649 trial, addition of nivolumab to chemotherapy had survival benefits as PD‐L1 CPS increased, with OS of 14.4 months for a PD‐L1 CPS of ≥ 5, 14.0 months for a PD‐L1 CPS of ≥ 1, and 13.8 months for all patients [10]. Similarly, in KEYNOTE‐859 and RATIONALE‐305, a higher PD‐L1 CPS was associated with favorable outcomes in patients treated with an anti‐PD‐1 antibody plus chemotherapy [11, 12]. However, our real‐world analysis showed that a high PD‐L1 CPS did not confer a survival advantage and, in fact, was associated with a trend toward poorer PFS. This finding suggests that the PD‐L1 CPS may have limited value as a predictive biomarker in clinical practice. Similarly, the G‐KNIGHT study, an observational study that evaluated the efficacy of first‐line Nivo‐CT for advanced or recurrent gastric cancer in Japan, found no significant difference in PFS according to the PD‐L1 CPS (7.5, 6.2, and 7.0 months for scores of < 1, 1 to < 5, and ≥ 5, respectively) [21]. Recent real‐world studies have also shown that, although outcomes may differ numerically across PD‐L1 CPS subgroups, PD‐L1 CPS alone is not a consistent prognostic factor in patients receiving first‐line immunotherapy‐based treatment for advanced gastric cancer [22, 23]. In the present study, PD‐L1 CPS was not centrally reviewed, and variations in specimen handling, such as fixation procedures and section thickness, may have introduced measurement error. Moreover, PD‐L1 CPS sums the percentages of both PD‐L1‐positive tumor and immune cells, leading to substantial interobserver variability [24]. These factors may have led to inconsistent classification of PD‐L1 CPS and prevented it from functioning as a reliable biomarker. In addition, the higher proportion of liver metastases in the PD‐L1 CPS ≥ 5 group (35.0% vs. 14.4%, p = 0.001) may have contributed to the relatively unfavorable outcomes observed in this group. Previous studies suggested that liver metastasis is an independent poor prognostic factor in advanced gastric cancer [25, 26]. In fact, the OS tended to be favorable in patients with PD‐L1 CPS ≥ 5 within the subgroup of patients without liver metastases. Furthermore, the small sample sizes of the lower‐CPS subgroups may have limited the statistical interpretation of comparisons between subgroups.
Considering that we did not compare Nivo‐CT with chemotherapy alone as a first‐line treatment, further discussion is needed regarding which patients truly benefit from first‐line Nivo‐CT. A PD‐L1 CPS subgroup analysis of CheckMate 649 found no significant difference in OS between Nivo‐CT and chemotherapy alone in patients with a PD‐L1 CPS of 1 to < 5 (HR 0.95, 95% CI 0.75–1.21) [27]. Likewise, a meta‐analysis of CheckMate 649, KEYNOTE‐859, and RATIONALE‐305 found that first‐line Nivo‐CT had limited survival benefit in patients with a PD‐L1 CPS of 1 to < 5 (median OS 12.1 vs. 12.0 months; HR 0.868, p = 0.018) [28]. In contrast, the findings of our present study were encouraging for patients with a PD‐L1 CPS of 1 to < 5, who had a median OS of 23.4 months and an ORR of 79.2%. These results suggest that first‐line Nivo‐CT may be preferable for this subgroup in clinical practice and are consistent with a previous finding that patients with a PD‐L1 CPS of 1 to < 5 may still benefit from ICI therapy [29].
Our multivariate analysis demonstrated that an ECOG PS ≥ 1, a GEJ primary, and ≥ 2 metastatic sites were independent poor prognostic factors. ECOG PS and the number of metastatic sites have consistently been identified as prognostic factors in advanced gastric cancer [21, 30, 31, 32], which is consistent with our present findings. On the other hand, several studies have found no difference in the prognosis according to whether the primary is a GEJ or gastric tumor in cases of unresectable advanced or recurrent disease [33, 34, 35]. The difference in median OS between GEJ and gastric primaries was minimal in CheckMate 649 (13.2 vs. 14.2 months) and ATTRACTION‐4 (14.9 vs. 16.6 months) [10, 19]. However, we found a statistically significant difference in median OS between GEJ and gastric primaries (12.3 vs. 20.0 months; HR 1.896, p = 0.028), although this finding should be interpreted with caution given that GEJ primaries accounted for only 24 patients (10.3%) of our cohort. To our knowledge, this is the first study to demonstrate that a GEJ primary is associated with a significantly poorer prognosis than a gastric primary in patients with HER2‐negative unresectable advanced or recurrent disease receiving first‐line ICI therapy plus chemotherapy. Differences in clinicopathological and molecular characteristics, such as Epstein–Barr virus infection status, genomic stability, and chromosomal instability, may contribute to the observed difference in the efficacy of Nivo‐CT between a GEJ primary and a gastric primary [34]. Studies comparing outcomes between HER2‐negative GEJ and gastric cancer treated with first‐line ICIs plus chemotherapy remain scarce, highlighting the need for further investigation.
MSI‐H/dMMR tumors were identified in 6.2% of patients evaluated for MSI status in the present study, and this proportion is consistent with previous reports [21, 36]. The lack of a significant survival improvement in patients with MSI‐H/dMMR tumors in comparison with those who had MSS/pMMR tumors is most likely attributable to the small number of cases with MSI‐H/dMMR. Notably, median OS and PFS were not yet reached and the ORR was 75.0%, indicating favorable outcomes in this subgroup. These findings suggest that patients with MSI‐H/dMMR tumors derive substantial benefit from first‐line Nivo‐CT, although the limited sample size warrants further investigation.
Grade ≥ 3 AEs attributable to Nivo‐CT were observed in 85 (36.3%) of our patients, whereas the rate was approximately 60% in both CheckMate 649 and ATTRACTION‐4 [10, 19]. However, our study was performed retrospectively, and the AE data may not have been fully captured. Furthermore, dose reductions at the start of treatment were permitted, and 50.0% of patients (n = 117) initiated Nivo‐CT with the dose of at least one drug reduced, which may also have contributed to the lower incidence of AEs.
This study has several limitations. First, it had a retrospective design, which might have introduced some selection bias. As all participating centers were located in Japan, the generalizability of our findings to other populations and healthcare systems may be limited. Second, we did not evaluate other biomarkers, such as Epstein–Barr virus infection status and claudin 18.2 expression in tumors, which may have provided further insights into molecular characterization and outcomes. Third, the objective response could be evaluated in only 43.2% of patients because those who had only nontarget lesions were excluded, which could introduce selection bias and potentially impact the generalizability of the findings.
In conclusion, this study found no significant difference in prognosis according to PD‐L1 CPS in patients who received first‐line Nivo‐CT for HER2‐negative unresectable advanced or recurrent advanced gastric or GEJ cancer in the real‐world setting. ECOG PS, a GEJ primary, and ≥ 2 metastatic sites were independent poor prognostic factors. However, further validation is required, particularly for patients with a GEJ primary. First‐line Nivo‐CT showed suggestive evidence of favorable outcomes in patients with MSI‐H/dMMR tumors.
Author Contributions
Naoki Takahashi: writing – review and editing, methodology, conceptualization. Kengo Nagashima: data curation, formal analysis, writing – review and editing, visualization. Hirokazu Shoji: supervision, conceptualization, methodology, writing – review and editing, project administration. Misa Ohnishi: investigation, writing – review and editing, data curation. Mitsuhiro Furuta: conceptualization, methodology, writing – review and editing. Rin Inamoto: investigation, writing – review and editing, data curation. Yosuke Kito: investigation, writing – review and editing, conceptualization, methodology, data curation. Nozomu Ogura: writing – original draft, conceptualization, investigation, validation, writing – review and editing, methodology, project administration, data curation.
Funding
The authors have nothing to report.
Ethics Statement
This study was approved by the ethics committees of all participating institutions (National Cancer Center Hospital, No. 2023‐266; Kanagawa Cancer Center, No. 2024‐29; Saitama Cancer Center, No. 1831; and Ishikawa Prefectural Central Hospital, No. 118).
Conflicts of Interest
Naoki Takahashi received honorarium from Ono Pharmaceutical Co. Ltd., Bristol‐Myers Squibb Co. Ltd., and Taiho Pharmaceutical Co. Ltd. Mitsuhiro Furuta received Honoraria from Astellas Pharmaceutical Co. Ltd., Bristol‐Myers Squibb Co. Ltd., Eli Lilly Co. Ltd., Merck Bio Pharma Co. Ltd., and Ono Pharmaceutical Co. Ltd. Hirokazu Shoji received honoraria from Astellas Pharmaceutical Co. Ltd., Abbvie, Astrazeneca, Elevation Oncology, Ono Pharmaceutical Co. Ltd., MSD K.K., Amgen K.K., and Takeda Pharmaceutical Co. Ltd. The other authors declare no conflicts of interest.
Supporting information
Figure S1: Survival outcomes in patients with microsatellite instability‐high or deficient mismatch repair. (A) Overall survival. (B) Progression‐free survival.
Table S1: Subsequent therapies.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Sung H., Ferlay J., Siegel R. L., et al., “Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries,” CA: A Cancer Journal for Clinicians 71, no. 3 (2021): 209–249. [DOI] [PubMed] [Google Scholar]
- 2. Lordick F., Kang Y. K., Chung H. C., et al., “Capecitabine and Cisplatin With or Without Cetuximab for Patients With Previously Untreated Advanced Gastric Cancer (EXPAND): A Randomised, Open‐Label Phase 3 Trial,” Lancet Oncology 14, no. 6 (2013): 490–499. [DOI] [PubMed] [Google Scholar]
- 3. Koizumi W., Narahara H., Hara T., et al., “S‐1 Plus Cisplatin Versus S‐1 Alone for First‐Line Treatment of Advanced Gastric Cancer (SPIRITS Trial): A Phase III Trial,” Lancet Oncology 9, no. 3 (2008): 215–221. [DOI] [PubMed] [Google Scholar]
- 4. Yamada Y., Higuchi K., Nishikawa K., et al., “Phase III Study Comparing Oxaliplatin Plus S‐1 With Cisplatin Plus S‐1 in Chemotherapy‐Naïve Patients With Advanced Gastric Cancer,” Annals of Oncology 26, no. 1 (2015): 141–148. [DOI] [PubMed] [Google Scholar]
- 5. Shah M. A., Bang Y. J., Lordick F., et al., “Effect of Fluorouracil, Leucovorin, and Oxaliplatin With or Without Onartuzumab in HER2‐Negative, MET‐Positive Gastroesophageal Adenocarcinoma: The METGastric Randomized Clinical Trial,” JAMA Oncology 3, no. 5 (2017): 620–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Guo L., Zhang H., Chen B., et al., “Nivolumab as Programmed Death‐1 (PD‐1) Inhibitor for Targeted Immunotherapy in Tumor,” Journal of Cancer 8, no. 3 (2017): 410–416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Kulangara K., Guerrero L., Posch A., et al., “Investigation of PD‐L1 Expression and Response to Pembrolizumab (Pembro) in Gastric Cancer (GC) and Cervical Cancer (CC) Using Combined Positive Score (CPS) and Tumor Proportion Score (TPS),” Journal of Clinical Oncology 36, no. 15 (2018): 4065. [Google Scholar]
- 8. Lei M., Siemers N., Pandya D., et al., “Association of PD‐L1 Combined Positive Score and Immune Gene Signatures With Efficacy of Nivolumab (NIVO) ± Ipilimumab (IPI) in Patients With Metastatic Gastroesophageal Cancer (mGEC),” Cancer Research 79 (2019): 2673. [Google Scholar]
- 9. Shitara K., Özgüroğlu M., Bang Y., et al., “Pembrolizumab Versus Paclitaxel for Previously Treated, Advanced Gastric or Gastro‐Oesophageal Junction Cancer (KEYNOTE‐061): A Randomised, Open‐Label, Controlled, Phase 3 Trial,” Lancet 392, no. 10142 (2018): 123–133. [DOI] [PubMed] [Google Scholar]
- 10. Janjigian Y. Y., Shitara K., Moehler M., et al., “First‐Line Nivolumab Plus Chemotherapy Versus Chemotherapy Alone for Advanced Gastric, Gastro‐Oesophageal Junction, and Oesophageal Adenocarcinoma (CheckMate 649): A Randomised, Open‐Label, Phase 3 Trial,” Lancet 398, no. 10294 (2021): 27–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Rha S. Y., Oh D. Y., Yañez P., et al., “Pembrolizumab Plus Chemotherapy Versus Placebo Plus Chemotherapy for HER2‐Negative Advanced Gastric Cancer (KEYNOTE‐859): A Multicentre, Randomised, Double‐Blind, Phase 3 Trial,” Lancet Oncology 24, no. 11 (2023): 1181–1195. [DOI] [PubMed] [Google Scholar]
- 12. Qiu M. Z., Oh D. Y., Kato K., et al., “Tislelizumab Plus Chemotherapy Versus Placebo Plus Chemotherapy as First‐Line Treatment for Advanced Gastric or Gastro‐Oesophageal Junction Adenocarcinoma (RATIONALE‐305): A Randomised, Double‐Blind, Phase 3 Trial,” BMJ 385 (2024): e078876. [DOI] [PubMed] [Google Scholar]
- 13. Shah M. A., Kennedy E. B., Alarcon‐Rozas A. E., et al., “Immunotherapy and Targeted Therapy for Advanced Gastroesophageal Cancer: ASCO Guideline,” Journal of Clinical Oncology 41, no. 7 (2023): 1470–1491. [DOI] [PubMed] [Google Scholar]
- 14. European Society for Medical Oncology (ESMO) , Gastric Cancer—ESMO Living Guideline, Version 1.4 (ESMO, 2024). [Google Scholar]
- 15. Ajani J. A., TA D'A., Bentrem D. J., et al., “Gastric Cancer, Version 2.2025, NCCN Clinical Practice Guidelines in Oncology,” Journal of the National Comprehensive Cancer Network 23, no. 5 (2025): 169–191. [DOI] [PubMed] [Google Scholar]
- 16. Shitara K., Fleitas T., Kawakami H., et al., “Pan‐Asian Adapted ESMO Clinical Practice Guidelines for the Diagnosis, Treatment and Follow‐Up of Patients With Gastric Cancer,” ESMO Open 9, no. 2 (2024): 102226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Le D. T., Uram J. N., Wang H., et al., “PD‐1 Blockade in Tumors With Mismatch‐Repair Deficiency,” New England Journal of Medicine 372, no. 26 (2015): 2509–2520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Le D. T., Durham J. N., Smith K. N., et al., “Mismatch Repair Deficiency Predicts Response of Solid Tumors to PD‐1 Blockade,” Science 357, no. 6349 (2017): 409–413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Kang Y. K., Chen L. T., Ryu M. H., et al., “Nivolumab Plus Chemotherapy Versus Placebo Plus Chemotherapy in Patients With HER2‐Negative, Untreated, Unresectable Advanced or Recurrent Gastric or Gastro‐Oesophageal Junction Cancer (ATTRACTION‐4): A Randomised, Multicentre, Double‐Blind, Placebo‐Controlled, Phase 3 Trial,” Lancet Oncology 23, no. 2 (2022): 234–247. [DOI] [PubMed] [Google Scholar]
- 20. Shimozaki K., Ooki A., Yoshino K., et al., “Investigating the Role of Immunotherapy for Real‐World Patients With HER2‐Negative Advanced Gastric Cancer Between 2011 and 2023,” Therapeutic Advances in Medical Oncology 17 (2025): 17588359251322670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Kadowaki S., Otsuka T., Minashi K., et al., “An Observational Study of the Effectiveness and Safety of Nivolumab Plus Chemotherapy for Untreated Advanced or Recurrent Gastric Cancer in Japanese Real‐World Settings: The G‐KNIGHT Study,” Gastric Cancer 28, no. 5 (2025): 955–967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mildanoglu M. M., Kutlu Y., Bas O., et al., “Prognostic and Predictive Value of Systemic Inflammatory Markers in Patients With Metastatic Gastric and GEJ Adenocarcinoma With PD‐L1 CPS Score ≥5: Turkish Oncology Group (TOG) Study,” Scientific Reports 15, no. 1 (2025): 25336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kutlu Y., Dae S. A., Yilmaz F., et al., “Real‐World Efficacy and Safety of First‐Line Nivolumab Plus Chemotherapy in Patients With Advanced Gastric, Gastroesophageal Junction, and Esophageal Adenocarcinoma: A Nationwide Observational Turkish Oncology Group (TOG) Study,” Cancers 16, no. 12 (2024): 2251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Robert M. E., Rushoff J., Jasani B., et al., “High Interobserver Variability Among Pathologists Using Combined Positive Score to Evaluate PD‐L1 Expression in Gastric, Gastroesophageal Junction, and Esophageal Adenocarcinoma,” Modern Pathology 36, no. 5 (2023): 100154. [DOI] [PubMed] [Google Scholar]
- 25. Chau I., Norman A. R., Cunningham D., et al., “Multivariate Prognostic Factor Analysis in Locally Advanced and Metastatic Esophago‐Gastric Cancer—Pooled Analysis From Three Multicentre, Randomized, Controlled Trials Using Individual Patient Data,” Journal of Clinical Oncology 22, no. 12 (2004): 2395–2403. [DOI] [PubMed] [Google Scholar]
- 26. Hagi T., Kurokawa Y., Kawabata R., et al., “Real‐World Effectiveness of Nivolumab for Previously Treated Unresectable or Recurrent Gastric Cancer: A Multicentre Retrospective Cohort Study,” British Journal of Cancer 123, no. 6 (2020): 965–972. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Zhao J. J., Yap D. W. T., Chan Y. H., et al., “Low Programmed Death‐Ligand 1‐Expressing Subgroup Outcomes of First‐Line Immune Checkpoint Inhibitors in Gastric or Esophageal Adenocarcinoma,” Journal of Clinical Oncology 40, no. 4 (2022): 392–402. [DOI] [PubMed] [Google Scholar]
- 28. Leone A. G., Mai A. S., Fong K. Y., et al., “Immune Checkpoint Inhibitors in Advanced Gastroesophageal Adenocarcinoma: A Series of Patient‐Level Meta‐Analyses in Different Programmed Death‐Ligand 1 Subgroups,” ESMO Open 9, no. 11 (2024): 103962. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Zhang X., Dai X., Liu A., et al., “Efficacy, Safety, and Biomarker Analysis of First‐Line Immune Checkpoint Inhibitors With Chemotherapy Versus Chemotherapy for Advanced Gastric Cancer: A Multicenter Retrospective Cohort Study,” BMC Medicine 22, no. 1 (2024): 585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Huo G., Liu W., Chen P., et al., “Efficacy of PD‐1/PD‐L1 Inhibitors in Gastric or Gastro‐Oesophageal Junction Cancer Based on Clinical Characteristics: A Meta‐Analysis,” BMC Cancer 23, no. 1 (2023): 143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Fanotto V., Cordio S., Pasquini G., et al., “Prognostic Factors in 868 Advanced Gastric Cancer Patients Treated With Second‐Line Chemotherapy in the Real World,” Gastric Cancer 20, no. 5 (2017): 825–833. [DOI] [PubMed] [Google Scholar]
- 32. Yamamoto S., Nakayama I., Takahari D., et al., “Current Prognostic Factors of Advanced Gastric Cancer Patients Treated With Chemotherapy: Real‐World Data From 12 Japanese Institutions,” Japanese Journal of Clinical Oncology 53, no. 10 (2023): 928–935. [DOI] [PubMed] [Google Scholar]
- 33. Kundel Y., Sternschuss M., Moore A., Perl G., Brenner B., and Goldvaser H., “Efficacy of Immune‐Checkpoint Inhibitors in Metastatic Gastric or Gastroesophageal Junction Adenocarcinoma by Patient Subgroups: A Systematic Review and Meta‐Analysis,” Cancer Medicine 9, no. 20 (2020): 7613–7625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Nakayama I., Takahari D., Wakatsuki T., et al., “Single‐Institute Comparison of the Efficacy of Systemic Chemotherapy for Oesophagogastric Junction Adenocarcinoma and Stomach Adenocarcinoma in a Metastatic Setting,” ESMO Open 5, no. 2 (2020): e000595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Pape M., Vissers P. A. J., Dijksterhuis W. P. M., et al., “Comparing Treatment and Outcomes in Advanced Esophageal, Gastroesophageal Junction, and Gastric Adenocarcinomas: A Population‐Based Study,” Therapeutic Advances in Medical Oncology 15 (2023): 17588359231162576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Komori A., Hironaka S., Kadowaki S., et al., “Prevalence and Clinicopathological Features of Microsatellite Instability‐High Metastatic or Recurrent Gastric and Esophagogastric Junction Cancer: WJOG13320GPS,” Gastric Cancer 28 (2025): 301–308. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1: Survival outcomes in patients with microsatellite instability‐high or deficient mismatch repair. (A) Overall survival. (B) Progression‐free survival.
Table S1: Subsequent therapies.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.