Abstract
Background:
Immune checkpoint inhibitors have shown promising results in the treatment of advanced gastric or gastroesophageal junction (G/GEJ) adenocarcinoma.
Objectives:
To evaluate the safety and efficacy of camrelizumab combined with S-1 plus oxaliplatin chemotherapy in the postoperative setting for pathological stage III (pStage III) G/GEJ adenocarcinoma.
Design:
This study was a prospective, multicenter, single-arm, phase II trial.
Methods:
Patients with pStage III G/GEJ adenocarcinoma were enrolled, receiving camrelizumab (200 mg), followed by oxaliplatin (130 mg/m2), both administered on day 1 of each 21-day cycle, and tegafur–gimeracil–oteracil potassium capsules (40–60 mg, twice daily) for 2 weeks, followed by a 7-day rest period. A total of eight treatment cycles were planned. The primary endpoints were the incidence of any-grade and grade 3 or 4 adverse events. The secondary endpoints were disease-free survival (DFS), overall survival (OS), and treatment completion rate.
Results:
Between September 2020 and March 2023, 52 patients were enrolled at three medical centers in Zhejiang Province, China. All 52 patients (100%) experienced treatment-related adverse events (TRAEs). Grade 3 or higher TRAEs were reported in 35 patients (67.3%). Events occurring in >10% of the patients included decreased neutrophil count, neutropenia, leukopenia, thrombocytopenia, and anemia. Eleven (21.2%) patients experienced TRAEs that led to the interruption or discontinuation of camrelizumab, including rash (1), hypothyroidism (1), hyperkalemia (1), interstitial pneumonia (2), cystitis or urethritis (2), and reactive cutaneous capillary endothelial proliferation (4). The actual 2-year DFS and OS rates were 82.4% and 86.3%, respectively, whereas the estimated 3-year DFS and OS rates were 69.1% and 71.3%, respectively.
Conclusion:
Camrelizumab combined with chemotherapy has a manageable and tolerable safety profile as an adjuvant treatment for pStage III G/GEJ adenocarcinoma. However, careful patient selection is necessary to identify patients most likely to benefit from combination therapy.
Trial registration:
ClinicalTrials.gov registry: NCT04515615; Date of registration: August 14, 2020; Weblink: https://clinicaltrials.gov/study/NCT04515615.
Keywords
Introduction
Gastric cancer (GC) is a public health concern worldwide, ranking fifth in both incidence and mortality, as reported by GLOBOCAN 2022. In 2022, there were 968,350 new cases and 659,853 deaths, of which nearly 40% occurred in China. 1 Despite a steady decline in GC incidence and mortality over the past half-century in most populations, the 5-year survival rate for patients with advanced-stage disease remains unsatisfactory, at approximately 20%–40%.2–4 Due to differences between Eastern and Western countries in the extent of standard lymphadenectomy—as well as survival benefits demonstrated in several phase III trials, including ACTS-GC, CLASSIC, RESOLVE, and JACCRO GC-075–9—the combination of surgery and adjuvant chemotherapy has become the standard treatment approach for pathological stage III (pStage III) gastric or gastroesophageal junction (G/GEJ) adenocarcinoma in Asian countries. However, the efficacy of this strategy is limited, with a 3-year disease-free survival (DFS) or recurrence-free survival (RFS) rate of approximately 60%.7,9 Improving the survival of these patients remains necessary, but challenging.
Immune checkpoint inhibitors (ICIs), particularly antibodies targeting programmed cell death protein-1 (PD-1) or its ligand PD-L1, have demonstrated notable efficacy in the treatment of unresectable advanced G/GEJ adenocarcinoma.10–12 In China, perioperative chemotherapy is recommended as the standard treatment for stage III G/GEJ adenocarcinoma, either in the neoadjuvant or adjuvant setting.7,13,14 Recently, the combination of ICIs with chemotherapy in the perioperative setting has gained considerable attention. These studies have shown improvements in pathological complete response (pCR) and major pathological response (MPR); however, survival outcomes remain pending.15–20 ATTRACTION-5, the first phase III clinical trial to evaluate adjuvant chemotherapy plus nivolumab in patients with pStage III G/GEJ adenocarcinoma, failed to demonstrate a significant benefit. The 3-year RFS rate was 68.4% in the nivolumab-plus-chemotherapy arm and 65.3% in the placebo-plus-chemotherapy arm (p = 0.44). 21 Thus, evidence supporting ICI-chemotherapy combinations in the adjuvant setting for this population remains limited.
Camrelizumab, a humanized, high-affinity, selective IgG4-κ monoclonal antibody that targets PD-1, has shown strong antitumor activity across a range of tumors.22–25 Unlike ATTRACTION-5, which used tegafur–gimeracil–oteracil (S-1) or capecitabine plus oxaliplatin, this study employed the S-1 plus oxaliplatin (SOX) regimen, which is widely used in China for advanced G/GEJ adenocarcinoma in neoadjuvant or adjuvant settings. 13 Here, we report the results of a phase II study to evaluate the safety and efficacy of camrelizumab plus SOX as adjuvant therapy in patients with pStage III G/GEJ adenocarcinoma.
Materials and methods
Study design and participants
We conducted a multicenter, single-arm, phase II trial of camrelizumab plus chemotherapy as adjuvant treatment for pStage III G/GEJ adenocarcinoma (FOCUS-02 trial). The study was registered at ClinicalTrials.gov (NCT04515615) on August 14, 2020, and was conducted across three medical centers in Zhejiang Province, China. The first patient was enrolled on September 9, 2020, and the last patient on March 10, 2023.
Key inclusion criteria required that patients had undergone D2 or D2 plus radical gastrectomy via open surgery (R0 resection) and were confirmed to have pStage III G/GEJ adenocarcinoma, according to the 8th edition of the AJCC/UICC TNM staging classification. Patients were enrolled regardless of PD-L1 expression. Additional inclusion and exclusion criteria are provided in Supplemental Text 1.
The study protocol was approved by the institutional review board and ethics committee of each center (latest version: V2.0, 2020.6.16). Written informed consent was obtained from all patients prior to enrollment, in accordance with the principles of the Declaration of Helsinki. This study conformed to the SPIRIT 2022 Outcomes checklist. 26
Interventions and assessments
Eligible patients initiated treatment within 4–6 weeks after surgery and were scheduled to receive intravenous camrelizumab 200 mg once on day 1, plus oxaliplatin 130 mg/m2 on day 1, and oral tegafur–gimeracil–oteracil potassium capsule as follows: 40 mg/dose for body surface area (BSA) <1.25 m2, 50 mg/dose for BSA ⩾1.25 m2 and <1.5 m2, and 60 mg/dose for BSA ⩾1.5 m2, administered twice daily on days 1–14. Each treatment cycle was repeated every 3 weeks, for a total of 8 cycles. Treatment was continued until disease recurrence, unacceptable toxicity, or withdrawal of consent. Dose modifications of camrelizumab were not permitted, and treatment could be interrupted or delayed for a maximum of 12 weeks during the trial intervention period. Dose reductions and interruptions were allowed for chemotherapy, with a maximum delay of 6 weeks.
Toxicities were closely monitored. Local laboratory assessments, including routine blood tests and blood biochemistry, were performed weekly during each cycle. Electrocardiogram (12-lead), echocardiography, and computed tomography (CT) scans of the lungs were conducted every two cycles or as clinically indicated. Adverse events (AEs) were monitored from the time of informed consent until 90 days after the final administration, and graded using the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. In case of grade 3 or higher hematologic or non-hematologic toxicity, the S-1 dose was reduced by one level (from 120 to 100 mg, from 100 to 80 mg, or from 80 to 50 mg), and the oxaliplatin dose was reduced from 130 to 100 mg/m2 or to 85 mg/m2.
Endpoints
The primary endpoint was safety, assessed by the incidence of AEs and grade 3 or 4 AEs. The secondary endpoints included DFS, overall survival (OS), and treatment completion rate. DFS was defined as the time from enrollment to disease recurrence or death from any cause, whichever occurred first. OS was defined as the time from enrollment to death from any cause. The treatment completion rate was defined as the proportion of patients who completed all eight cycles of the scheduled regimen.
Statistical analysis
Categorical variables were summarized as frequency counts and percentages, while continuous variables were reported as medians and ranges. The primary outcome, AE-related toxicity, was reported by category and grade. The Kaplan–Meier method was used to estimate DFS and OS. All statistical analyses were performed using SPSS software (version 25.0; SPSS Inc., Chicago, IL, USA), with a significance level set at 5%.
Results
Patients’ characteristics
From September 2020 to March 2023, 57 patients from three sites were assessed for eligibility, and 52 patients were enrolled to receive camrelizumab and chemotherapy (Figure 1). The median patient age was 60 (range, 35–75) years. Forty patients (76.9%) were men and 12 (23.1%) were women. The location of the primary tumor was as follows: 8 patients (15.4%) had tumors in the GEJ, 16 (30.8%) had tumors in the gastric corpus, 2 (3.8%) had tumors in both the gastric corpus and antrum, and 26 (50%) had tumors in the gastric angle, antrum, or pylorus. The pathological stage was IIIA in 22 patients (42.3%), IIIB in 19 (36.5%), and IIIC in 11 (21.2%). According to the Lauren classification, 10 patients (19.2%) had intestinal-type tumors, 16 (30.8%) had diffuse-type, and 22 (42.3%) had mixed-type histology; data were not available for 4 patients (7.7%). Pathologically, 51 patients (98.1%) had poorly differentiated tumors, and 1 patient (1.9%) had a moderately differentiated tumor. Lymphovascular or perineural invasion was observed in 49 patients (94.3%). The demographic and baseline clinical characteristics are summarized in Table 1.

CONSORT flow diagram of the study population.
Demographic and clinical characteristics.
dMMR, deficient mismatch repair; EBER, epstein-barr virus encoded small RNA; ECOG, Eastern Cooperative Oncology Group; GEJ, gastroesophageal junction; pMMR, proficient mismatch repair.
Exposure and safety
All 52 patients received at least one dose of the study drug and were included in the safety analysis. A total of 22 patients (42.3%) completed the assigned regimen for 8 cycles; 30 (57.7%) completed 7 or more cycles, and 38 (73.1%) completed 6 or more cycles. Treatment-related adverse events (TRAEs) led to chemotherapy dose reductions in 29 patients (55.8%; Table 2). Thirty of the 52 patients (57.7%) discontinued the assigned regimen for the following reasons: disease progression (1 patient, 1.9%), AEs (15 patients, 28.8%), patient decision (7 patients, 13.5%), investigator decision (3 patients, 5.8%), and the COVID-19 pandemic (4 patients, 7.7%; Figure 1). All 52 patients (100%) experienced TRAEs. The most common TRAEs of any grade are fatigue, decreased white blood cell and neutrophil counts, decreased appetite, and reactive cutaneous capillary endothelial cell proliferation (RCCEP). Grade ⩾ 3 TRAEs were reported in 35 (67.3%) patients. Events occurring in >10% of the patients included decreased neutrophil count (n = 26, 50%), decreased white blood cell count (n = 7, 13.5%), decreased platelet count (n = 7, 13.5%), and anemia (n = 7, 13.5%). Treatment-related serious AEs occurred in 6 patients (11.5%), and no treatment-related deaths were reported. Eleven patients (21.2%) experienced TRAEs that led to the interruption or discontinuation of camrelizumab, including rash (1 patient), hypothyroidism (1 patient), hyperkalemia (1 patient), interstitial pneumonia (2 patients), cystitis or urethritis (2 patients), and RCCEP (4 patients; Table 3). No unexpected toxicities were observed. Furthermore, no significant associations were found between age and the severity or pattern of TRAEs (Supplemental Table 1).
Exposure and disposition.
TRAE, treatment-related adverse events.
Treatment-related adverse events (N = 52).
RCCEP, reactive cutaneous capillary endothelial proliferation.
Survival
As of the data cutoff on April 8, 2025, one patient was lost to follow-up. The median follow-up duration (time from enrollment to the last contact) was 35 (range, 6–55) months. Fifteen patients (28.8%) experienced cancer recurrence, and 12 (23%) died. The sites of first relapse were the peritoneum (nine patients), lymph nodes (three patients), hematogenous sites (four patients), and local recurrence (two patients), with some patients experiencing relapse at multiple sites. The 2-year DFS and OS rates were 82.4% and 86.3%, respectively. The estimated 3-year DFS and OS rates were 69.1% and 71.3%, respectively. However, the actual 3-year survival rates had not yet been reached at the time of analysis (Figure 2(a) and (b)). The subgroup survival analysis is presented in Supplemental Table 2.

Disease-free survival and overall survival. Disease-free survival (range: 4–55 months, (a) and overall survival (range: 6–55 months, (b) with the median follow-up of 35 months.
Discussion
Over the past decade, the use of ICIs has provided clinical benefits and improved survival outcomes in patients with advanced G/GEJ adenocarcinoma.27,28 However, evidence supporting the use of adjuvant chemotherapy plus ICIs in patients with locally advanced, resectable G/GEJ adenocarcinoma after curative gastrectomy remains limited.
In our study, the adjuvant chemotherapy regimen differed from that used in the ATTRACTION-5 study, which is commonly used in China according to the results of the RESOLVE study,7,14 demonstrating that adjuvant SOX is non-inferior to adjuvant CapeOx in patients with locally advanced G/GEJ adenocarcinoma who underwent D2 gastrectomy. Several studies have demonstrated that camrelizumab combined with SOX has a promising efficacy and an acceptable safety profile in GC, with higher objective response rates and prolonged survival in unresectable late-stage GC, or higher pCR and MPR rates in resectable locally advanced GC.24,29,30 Our study focused on patients who underwent curative resection for pStage III G/GEJ adenocarcinoma. In this cohort, 35 patients (67.3%) experienced grade ⩾3 TRAEs, which was higher than previously reported in neoadjuvant or metastatic settings.24,30 This may be attributed to worse performance status and potential nutritional risk in the postoperative population. RCCEP has been reported as the most common AE associated with camrelizumab, with an incidence ranging from 34.5% to 72%,23,24,30 and only a small proportion of cases being grade ⩾3. In our study, 43 patients (82.7%) experienced RCCEP, all of which were graded as 1 (n = 30; 57.7%) or 2 (n = 13; 25.0%). Immune-related AEs involving the urinary tract and bladder are rarely reported.31,32 In our study, two patients (3.8%) were suspected of having urethritis/cystitis, although renal biopsy was not performed. Diagnostic evaluations included urinalysis, renal function testing, and CT to differentiate among bacterial cystitis, metastasis, and chemotherapy-related cystitis. Camrelizumab was discontinued, and the symptoms improved with continued chemotherapy alone. Overall, camrelizumab combined with SOX was well tolerated in our study, and no treatment-related deaths were reported. However, careful management of AEs remains essential.
In our study, 22 patients (42.3%) completed 8 cycles, and 38 (73.1%) completed 6 or more cycles. The secondary endpoint treatment completion rate was comparable to that reported in other trials. In the ACTS-GC study, 65.8% of patients completed 1 year of S-1 treatment. In the CLASSIC study, 67% of patients completed eight cycles of CapeOx. In the JACCRO GC-07 study, which included patients with pStage III GC, 67% of patients completed all six cycles of docetaxel. In the RESOLVE study, 40.9% completed at least six cycles of CapeOx and 22.3% completed all eight cycles; for adjuvant SOX, the corresponding rates were 35.6% and 19.4%, respectively. The ATTRACTION-5 study suggested that patients with tumor-cell PD-L1 expression ⩾1%, ECOG performance status score of 1, and pStage IIIC G/GEJ adenocarcinoma may benefit more from adjuvant chemotherapy plus nivolumab. However, due to our single-arm design, we were unable to identify the patient subgroups most likely to benefit. Nevertheless, our data suggest that worse DFS was associated with a higher ECOG performance score (1 vs 0; HR = 2.752, 95% CI: 0.997–7.592, p = 0.051) and more advanced pathological stage (pStage IIIB/IIIC vs pStage IIIA; HR = 3.819, 95% CI: 1.074–13.575, p = 0.038), while worse OS was associated with advanced age (⩾65 vs <65 years; HR = 3.247, 95% CI: 1.030–10.237, p = 0.044). The actual 2-year DFS and OS rates were 82.4% and 86.3%, respectively, whereas the estimated 3-year DFS and OS rates were 69.1% and 71.3%, respectively. In the ARTIST 2 trial, the estimated 3-year DFS was 74.3% in patients with D2-resected, stage II or III, node-positive GC receiving adjuvant SOX. 33 The RESOLVE study reported a 3-year DFS of 56.5% in the adjuvant SOX group for clinical Stage III G/GEJ adenocarcinoma. 7 In addition, the ATTRACTION-5 reported a 3-year RFS of 65.3% in pStage III G/GEJ adenocarcinoma treated with adjuvant regimens of S-1 or CapeOx. 21 These comparisons suggest that camrelizumab plus SOX may offer a potential advantage in 3-year DFS over SOX alone in pStage III G/GEJ adenocarcinoma; however, the actual 3-year DFS and OS outcomes remain to be confirmed.
This study had some limitations. First, the single-arm design, small sample size, and lack of a control group limit the generalizability of the results. Second, the completion rate of postoperative treatment was relatively low (42.3%), partly due to the COVID-19 pandemic, which may have introduced bias. Third, biomarker analysis for predicting AEs, efficacy, and prognosis was not conducted. Previous studies have reported circulating tumor DNA (ctDNA) as a promising biomarker for predicting relapse in solid tumors. For instance, ctDNA positivity after adjuvant chemotherapy has been associated with significantly increased recurrence risk in stage II/III resectable GC (positive vs negative; HR = 14.99, 95% CI = 3.08–72.96, p < 0.001). 34 Future studies should integrate ctDNA, radiomics, and other pathological biomarkers.
Conclusion
In conclusion, this study suggests that camrelizumab combined with chemotherapy has a manageable and tolerable safety profile as an adjuvant treatment for pStage III G/GEJ adenocarcinoma. Nevertheless, identifying patients who are most likely to benefit from this combination therapy remains a key objective, and future research should prioritize biomarker-driven patient selection.
Supplemental Material
sj-docx-1-tam-10.1177_17588359251355781 – Supplemental material for Camrelizumab plus SOX chemotherapy as adjuvant therapy for pathological stage III gastric or gastroesophageal junction adenocarcinoma: a prospective, multicenter, single-arm, phase II trial
Supplemental material, sj-docx-1-tam-10.1177_17588359251355781 for Camrelizumab plus SOX chemotherapy as adjuvant therapy for pathological stage III gastric or gastroesophageal junction adenocarcinoma: a prospective, multicenter, single-arm, phase II trial by Hailong Jin, Jiren Yu, Yuan Gao, Xiaosun Liu, Hongtao Xu, Haiping Jiang, Shenkang Zhou, Nong Xu, Yulong Zheng, Qianyun Shen, Zhicheng Zhao, Dan Wu, Qing Zhang and Kankai Zhu in Therapeutic Advances in Medical Oncology
Footnotes
Acknowledgements
This is an investigator-initiated trial. The authors would like to thank all of the physicians, nurses, researchers, and patients for their great cooperation and valuable contribution in this study.
Declarations
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
