Real-World Efficacy of Cadonilimab Plus Chemotherapy as Second-Line Therapy in Immunotherapy-Pretreated Advanced HER2 Negative Gastric or Gastroesophageal Junction Cancer

Introduction

Gastric or gastroesophageal junction cancer (G/GEJC) ranks fifth in global tumor incidence and represents the fifth leading cause of cancer-related mortality, imposing a significant health burden worldwide.¹ In China, G/GEJC remains the third leading cause of cancer-related deaths, accounting for approximately 44% of global cases. Among these, only approximately 20% are diagnosed at an early stage.^2–4 Patients with G/GEJC still require more effective treatment options.

In recent years, immunotherapy has emerged as a promising therapeutic approach for various cancers, including G/GEJC. Phase III clinical trials including CheckMate-649,⁵ RATIONALE-305⁶ and KEYNOTE-859⁷ have established immunotherapy combined with chemotherapy as the standard first-line treatment for advanced HER2-negative gastric cancer. However, existing second-line standards (eg, REGARD⁸ and RAINBOW⁹ studies), which enrolled patients progressing after first-line chemotherapy, demonstrate limited efficacy. Crucially, there remains no established standard treatment for patients progressing after first-line immunotherapy - a critical unmet clinical need requiring urgent resolution.

Cadonilimab (AK104), developed by Akeso Biopharma, is a bispecific antibody targeting both PD-1 and CTLA-4. It is designed to maintain potent antitumor activity while reducing adverse events (AEs) related to autoimmunity and inflammation. The agent has demonstrated clinically meaningful benefits in cervical cancer and hepatocellular carcinoma.^10,11 In previous studies, cadonilimab combined with chemotherapy has demonstrated promising efficacy as first-line treatment for advanced G/GEJC.^12,13 Based on the positive results from the COMPASSION-15 study, cadonilimab has been approved in China for the first-line treatment of advanced G/GEJC.¹³ The AK109-201 study demonstrated that cadonilimab combined with pulocimab (an anti-VEGFR-2 mAb), and paclitaxel as second-line therapy provided significant survival benefits for patients who had progressed on prior immunotherapy, highlighting its substantial clinical potential. A phase III trial is currently underway to validate these findings.¹⁴ The biological rationale for evaluating cadonilimab after prior immunotherapy failure hinges on the potential for compensatory upregulation of alternative checkpoints like CTLA-4, leading to a multi-faceted T-cell exhaustion. We hypothesized that simultaneously targeting both PD-1 and CTLA-4, two inhibitory receptors on T cells, could more effectively reverse this exhaustion and overcome adaptive resistance, providing a strong rationale for immune rechallenge in the secondline setting.^15,16 Therefore, in this study, we retrospectively investigated the efficacy and safety of cadonilimab combined with chemotherapy as a second-line treatment for advanced G/GEJC patients previously treated with immunotherapy. Additionally, we analyzed potential factors influencing the efficacy of cadonilimab in the second-line setting after prior immunotherapy failure.

Methods

Patients

This retrospective study included patients with advanced G/GEJC who experienced disease progression after first-line immunotherapy at the Affiliated Yantai Yuhuangding Hospital of Qingdao University between October 2022 and April 2025. This study was approved by the Clinical Research Ethics Committee of the Affiliated Yantai Yuhuangding Hospital of Qingdao University and conducted in accordance with the Declaration of Helsinki (YYYIRB-IIT [2025]014). As this was a retrospective study, the requirement for informed consent was waived.

A total of 112 patients were enrolled in this study. Among them, 50 patients received second-line treatment with cadonilimab combined with albumin-bound paclitaxel/irinotecan (Cohort A), while 62 patients received albumin-bound paclitaxel/irinotecan as monotherapy (Cohort B). The inclusion criteria were as follows: age ≥ 18 years, Eastern Cooperative Oncology Group (ECOG) performance status ≤ 2, prior treatment with immune checkpoint inhibitors, prior progression on first-line immunotherapy, HER2 negative and measurable lesions according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. The exclusion criteria included prior receipt of any anti-tumor therapy other than neoadjuvant, adjuvant, or first-line treatment, as well as the presence of serious or uncontrolled systemic diseases.

Treatments

Patients who met the inclusion criteria and had received either chemotherapy alone or chemotherapy combined with cadonilimab (10 mg/kg, every three weeks (Q3W)) were enrolled in this study. Chemotherapeutic agents included nab-paclitaxel (260 mg/m², Q3W), paclitaxel (175 mg/m², Q3W), and docetaxel (75 mg/m², Q3W). After six to eight treatment cycles, patients underwent cadonilimab monotherapy as maintenance treatment until either disease progression or unacceptable toxicity occurred.

Assessments

The clinical response was assessed every two cycles according to RECIST 1.1. Treatment responses were categorized as complete response (CR), partial response (PR), stable disease (SD), or progressive disease (PD). The primary endpoints were progression-free survival (PFS) and overall survival (OS). PFS was defined as the time from the initiation of second-line treatment to disease progression or patient death, while OS was defined as the time from the initiation of second-line treatment to death from any cause. The secondary endpoints included objective response rate (ORR), disease control rate (DCR), and safety. ORR was defined as the proportion of patients assessed as PR or CR, and DCR was defined as the proportion of patients assessed as PR, CR, or SD. AEs were evaluated according to the Common Terminology Criteria for Adverse Events (CTCAE) version 4.0.

Statistical Analysis

All data were analyzed using SPSS version 27.0 (IBM, New York, USA). All tests were two-sided, with a significance level set at p < 0.05. Categorical variables were compared using the Chi-square test or Fisher’s exact test, as appropriate. The Fisher’s exact test was used if the expected frequency in any cross-tabulation cell was below 5. PFS and OS for Cohort A and Cohort B were calculated using the Kaplan-Meier method, with results expressed as medians and 95% confidence intervals (CIs). To identify factors influencing efficacy, survival analysis between subgroups of the two cohorts was performed using the Log rank test. Hazard ratios (HRs) were calculated through univariate Cox proportional hazards regression. Multivariate Cox regression analysis was performed to adjust for potential confounders and evaluate the impact of clinical factors on survival outcomes.

Result

Patient Characteristics

From October 2022 to April 2025, a total of 112 patients were enrolled in this study. Among them, 50 patients received cadonilimab combined with chemotherapy, designated as Cohort A, while 62 patients received chemotherapy alone, designated as Cohort B. The median follow-up durations for Cohort A and Cohort B were 8.7 months and 7.3 months, respectively. Patients in both cohorts were comparable in terms of age, sex, ECOG performance status, and other baseline characteristics (Table 1). The prior first-line immunotherapy regimens included sintilimab (100/112, 89.3%), nivolumab (5/112, 4.5%), tislelizumab (5/112, 4.5%), and pembrolizumab (2/112, 1.8%). The recruitment process for our study is illustrated in Figure 1. A total of 18 patients (10 in Cohort A and 8 in Cohort B) discontinued second-line treatment due to AEs, poor health conditions, or personal reasons. The median number of treatment cycles was 6.

Table 1 Baseline Characteristics of Patients

Figure 1 Flowchart of the recruitment process.

Efficacy and Prognostic Factors

As of the follow-up cutoff date (April 30, 2025), the mPFS was 4.9 months (95% CI: 3.9–6.0) in Cohort A compared with 3.8 months (95% CI: 2.8–4.8) in Cohort B (Figure 2A). For overall survival, Cohort A demonstrated a mOS of 10.3 months (95% CI: 8.8–11.8), while Cohort B showed an mOS of 7.4 months (95% CI: 6.9–7.9) (Figure 2B). Among the 50 patients in Cohort A, 17 (34.0%) achieved partial response (PR), 20 (40.0%) had stable disease (SD), and 13 (26.0%) experienced progressive disease (PD). In Cohort B, which included 62 patients, 11 (17.7%) achieved PR, 23 (37.1%) had SD, and 28 (45.2%) showed PD. The overall response rate (ORR) in Cohort A was 34.0%, significantly higher than that in Cohort B (17.7%; P = 0.048). Similarly, the disease control rate (DCR) in Cohort A was 74.0%, which was also significantly higher than that in Cohort B (54.8%; P = 0.036) (Table 2).

Table 2 Treatment Response Outcome

Figure 2 (A) PFS in cohort A versus cohort B, (B) OS in cohort A versus cohort B.

Subgroup analyses were performed to identify factors associated with survival outcomes. Initial univariate analysis using Kaplan-Meier method and Log rank test revealed that neither age, sex, Ki-67, nor histological grade were significantly associated with patient survival. Notably, significant benefits in PFS and/or OS were observed in several key subgroups treated with cadonilimab combination therapy. Specifically, patients with fewer than 3 metastatic sites showed significantly improved PFS (5.8 vs 4.8 months; HR=2.0, 95% CI: 1.1–3.6, p=0.015), while patients with liver metastases demonstrated superior OS (10.3 vs 7.2 months; HR=2.7, 95% CI: 1.4–4.3, p=0.011). The most pronounced benefit was observed in patients with PD-L1 CPS ≥1, who achieved significantly longer PFS (5.8 vs 3.8 months; HR=2.4, 95% CI: 1.4–4.3, p=0.001) and OS (11.5 vs 7.4 months; HR=2.9, 95% CI: 1.5–5.4, p<0.001) (Table 3, Figures 3 and 4).

Table 3 Summary of mPFS, mOS, and HRs for Cohorts A and B

Figure 3 Subgroup analysis of PFS. (A) Number of metastatic sites <3, (B) PD-L1 CPS≥1.

Figure 4 Subgroup analysis of OS. (A) Liver metastasis, (B) PD-L1 CPS≥1.

To adjust for potential confounders, these significant variables were subsequently included in a multivariate Cox regression model. The results confirmed that the number of metastatic sites, the presence of liver metastases, and PD-L1 CPS status remained independent prognostic factors for survival, retaining their statistical significance after multivariate adjustment (Table 4 and 5). Subgroups for ECOG performance status and primary tumor location were not analyzed due to insufficient sample size in certain categories.

Table 4 Multivariable Cox Regression Analysis for PFS

Table 5 Multivariable Cox Regression Analysis for OS

Safety

Table 6 summarizes the treatment-related AEs and immune-related AEs observed during the study period. In Cohort A, the most common treatment-related AEs were decreased neutrophil count (20/50, 40%) and anemia (18/50, 36.0%). The most frequent grade 3–4 AEs were abnormal liver function (6/50, 12.0%) and decreased neutrophil count (5/50, 10.0%). In Cohort B, the most common treatment-related AEs were anemia (18/62, 29.0%) and decreased neutrophil count (16/62, 25.8%). The most frequent grade 3–4 AEs in this cohort were decreased neutrophil count (4/62, 6.5%) and anemia (4/62, 6.5%). The p-values presented in Table 6 are used to indicate whether there were statistically significant differences in the incidence of each treatment-related AE between the two cohorts. There were no statistically significant differences in the incidence of treatment-related AEs between the two cohorts, and no treatment-related deaths were observed.

Table 6 Treatment-Related AEs Occurred During treatments

Discussion

Our study demonstrated statistically significant differences in PFS and OS between the two treatment cohorts in the overall population (PFS: 4.9 months vs 3.8 months, HR = 1.6, 95% CI 1.0–2.4, p = 0.024; OS: 10.3 months vs 7.4 months, HR = 1.5, 95% CI 1.0–2.4, p = 0.046). Additionally, notable differences were found in ORR (34.0% vs 17.7%, p = 0.048) and DCR (74.0% vs 54.8%, p = 0.036), demonstrating that cadonilimab exhibited meaningful efficacy with an acceptable safety profile in immunotherapy-pretreated advanced G/GEJC patients.

The standard second-line treatment for advanced G/GEJC consists of chemotherapy alone or ramucirumab (an anti-VEGFR-2 monoclonal antibody) combined with chemotherapy.^17,18 The RAINBOW trial established ramucirumab as the standard second-line therapy in 2014;⁹ however, current clinical accessibility remains limited. In recent years, the rapid advancement of immunotherapy has made it a cornerstone in the treatment of advanced G/GEJC. Immune checkpoint inhibitors such as pembrolizumab, cadonilimab, tislelizumab, nivolumab, and sintilimab have demonstrated robust efficacy in first-line settings,^{5–7,13,19,20} transforming the second-line patient population from previously “chemotherapy-pretreated” to “immunotherapy plus chemotherapy-pretreated.” The optimal therapeutic strategy for immunotherapy-experienced patients following resistance to first-line immunotherapy remains a critical unmet clinical challenge that warrants further investigation. The bispecific antibody cadonilimab, targeting both PD-1 and CTLA-4, offers a novel therapeutic option. Our study evaluated the efficacy and safety of cadonilimab combined with chemotherapy as an immunotherapy rechallenge strategy for second-line G/GEJC patients.

In the RAINBOW trial,⁹ ramucirumab plus paclitaxel demonstrated statistically significant improvements in both PFS (4.4 vs 2.9 months; HR=0.635, 95% CI: 0.536–0.752, p<0.0001) and OS (9.6 vs 7.4 months; HR=0.807, 95% CI: 0.678–0.962, p=0.0169) compared to placebo plus paclitaxel. The FRUTIGA study²¹ showed fruquintinib combined with paclitaxel significantly enhanced PFS (5.6 vs 2.7 months; HR=0.57, p<0.0001), ORR (42.5% vs 22.4%, p<0.0001), and DCR (77.2% vs 56.3%, p<0.0001) versus placebo-paclitaxel in the second-line setting, although no OS benefit was observed. Notably, in the immunotherapy-pretreated subgroup of FRUTIGA, fruquintinib achieved markedly prolonged PFS (6.4 vs 1.8 months; HR=0.38, p=0.0003) without reaching statistical significance in OS. While the RAINBOW study demonstrated favorable outcomes, its study population was limited to patients who had failed first-line chemotherapy. Similarly, the FRUTIGA trial showed PFS benefits but no OS advantage in immunotherapy-pretreated patients. In contrast, our study specifically enrolled patients with progression after first-line immunotherapy and revealed that second-line cadonilimab combined with chemotherapy achieved statistically significant improvements in both PFS and OS compared to chemotherapy alone.

Our findings are further corroborated by emerging evidence on dual-checkpoint inhibition in this challenging setting. AK109-201 study evaluated cadonilimab combined with pulocimab and paclitaxel in G/GEJC patients who failed prior immunotherapy. The regimen demonstrated promising efficacy, including an ORR of 48.0%, DCR of 96.0%, and a mPFS of 6.8 months. Together, these data solidify the role of cadonilimab based therapy as a potent strategy for immunotherapy rechallenge.¹⁴

In the PD-L1 CPS ≥1 subgroup, patients treated with cadonilimab demonstrated significantly superior PFS and OS compared to chemotherapy alone, suggesting that cadonilimab effectively overcomes immunotherapy resistance in previously treated patients and exhibits robust efficacy in PD-L1 CPS ≥1 advanced G/GEJC. Our findings suggest that the efficacy of cadonilimab in this immunotherapy-pretreated population may be attributed to its ability to overcome resistance via dual-blockade of the compensatory PD-1 and CTLA-4 pathways. In gastric cancer, high expression of PD-L1 and CTLA-4 is associated with poor prognosis.²² Tumor cells evade immune surveillance by upregulating CTLA-4 and PD-1 on T lymphocytes.^23,24 Anti-PD-1 antibodies promote the proliferation and activation of activated and exhausted T cells, enhancing T cell-mediated cytotoxicity. Anti-CTLA-4 antibodies enhance the priming and activation of naïve T cells, driving early expansion of the T cell receptor repertoire, though this may disrupt immune tolerance in normal tissues.¹⁵ The synergistic action of these two mechanisms is postulated to effectively reverse T cell exhaustion and overcome immune resistance. While the two mechanisms can synergize for enhanced efficacy, they may also increase toxicity. Cadonilimab is a tetravalent PD-1/CTLA-4 bispecific antibody that overcomes resistance to PD-1 or CTLA-4 monotherapy through dual-pathway blockade, effectively reactivating exhausted T cells co-expressing PD-1 and CTLA-4. This agent enhances CD8+ T cell activation capability even in low-antigen-density microenvironments. Its Fc-null design eliminates Fc-mediated effector functions: antibody-dependent cellular cytotoxicity (ADCC), antibody-dependent cellular phagocytosis (ADCP), and complement-dependent cytotoxicity (CDC), thereby significantly reducing the incidence of severe adverse events. Cadonilimab exhibits enhanced binding avidity in PD-1/CTLA-4 high-density regions, enabling preferential enrichment within the tumor microenvironment—thereby maintaining therapeutic efficacy while reducing systemic toxicity.²⁵ In PD-L1 CPS<1 patients, neither PFS nor OS demonstrated significant improvement, highlighting the unmet need for novel therapeutic approaches in this PD-L1-negative population as second-line treatment.

Notably, cadonilimab-pretreated patients with fewer than 3 metastatic sites showed superior PFS compared to controls (5.8 vs 4.8 months; HR=2.0, 95% CI: 1.1–3.6, p=0.015). This finding suggests that patients with fewer metastatic lesions may benefit from enhanced immunotherapeutic efficacy of cadonilimab, potentially attributable to lower tumor burden and more complete immune function. This observation aligns with the report by Wei et al, which demonstrated improved immunotherapy outcomes in patients with fewer tumor lesions and lower disease burden.²⁶ Particularly striking benefits were observed in liver metastasis patients, where cadonilimab significantly improved OS (10.3 vs 7.2 months; HR=2.7, 95% CI: 1.4–4.3, p=0.011). Liver metastasis typically indicates poorer prognosis and diminished response to immunotherapy, likely attributable to the enrichment of immunosuppressive cells within hepatic metastatic lesions.^27,28 The cancer cells invading the liver may activate liver-specific immune tolerance mechanisms, thereby compromising systemic antitumor immune responses and reducing the efficacy of immunotherapy.²⁹ Research evidence indicates that CTLA-4/PD-1 bispecific antibodies may potentially enhance immunotherapeutic efficacy in liver metastasis patients more effectively than PD-1 monotherapy through coordinated modulation of Tregs, MDSCs, and effector T cells.²⁹ Besides, the COMPASSION-15 study similarly demonstrated enhanced clinical benefits in the liver metastases subgroup.¹³ Cadonilimab’s dual immune checkpoint inhibition may represent a promising therapeutic strategy to overcome treatment resistance in this traditionally refractory patient population.

Looking forward, our results, alongside the AK109-201 trial, support cadonilimab as a viable option for immunotherapy pretreated patients. Future efforts should focus on validating these findings in larger trials and identifying predictive biomarkers beyond PD-L1. Given drug accessibility constraints in some regions, exploring combinations of cadonilimab with locally available agents like ramucirumab or fruquintinib, alongside chemotherapy, represents a critical and practical research direction.

Our study has several limitations. First, as a single-center retrospective analysis, the limited sample size may not fully represent treatment outcomes in patients from other institutions. Other variables such as tumor mutational burden could influence survival outcomes; however, our retrospective design precluded comprehensive data acquisition for such analyses. The limited sample sizes in certain subgroups may also compromise the statistical reliability of our findings. These constraints prompt us to consider conducting a prospective multicenter study to obtain larger and more comprehensive sample data, thereby further validating our conclusions. Furthermore, the precise mechanisms by which cadonilimab overcomes immunotherapy resistance remain to be fully elucidated and warrant further investigation.

Conclusion

In the second-line treatment for immunotherapy-pretreated G/GEJC patients, the combination of cadonilimab and chemotherapy has been shown to achieve significantly better outcomes in terms of PFS, OS, ORR, and DCR compared to chemotherapy alone in the overall population. In particular, patients with PD-L1 CPS ≥1 demonstrated superior survival outcomes, achieving statistically significant improvements in both PFS and OS, whereas no benefit was seen in patients with CPS <1. Notably, cadonilimab demonstrated differential clinical benefits based on metastatic patterns: patients with fewer than 3 metastatic sites achieved superior PFS, while those with liver metastases exhibited particularly enhanced OS outcomes. These results suggest that cadonilimab plus chemotherapy may represent a promising therapeutic option for immunotherapy rechallenge in second-line G/GEJC.