Gastric cancer (GC) is the fourth most common cause of cancer-related death worldwide, accounting for nearly 800,000 deaths in 2020; approximately 50% of these deaths (an estimated 373,789) occurred in China (1). The poor prognosis of GC is related to local recurrence, gross peritoneal dissemination, invasion of other organs, and extensive distant metastasis (2). With current treatment strategies, patients with localised disease usually have a good prognosis, but metastatic cancers are generally incurable because of their spread to distant locations, leaving patients with no chance of radical resection and only conservative medical treatment to control the progression of the disease (3). The median survival of Chinese patients with metastatic GC varies from 3.9 months to 18.4 months based on the metastatic location (4–8). Therefore, identifying the optimal therapies for patients with metastatic GC is still needed.

The liver is a frequent site of distant metastasis in GC, with an incidence of 5–34% (9, 10). Despite significant efforts to improve survival of patients with liver metastasis (LM), the prognosis remains poor. Therapeutic surgery provides a potential cure for liver metastasis from gastric cancer (LMGC); however, surgery is currently not a standard treatment option for patients with advanced GC other than palliative surgery for bleeding, obstruction, or perforation caused by the tumour. Therefore, palliative chemotherapy is still regarded as the standard treatment modality for patients with metastatic GC (11–13). The SOPP trial showed that S-1 plus oxaliplatin (SOX) regimen can be recommended as a first-line treatment for metastatic or recurrent GC, and the median OS can be extended to12.9 months (14). According to recent studies (15–17), the median survival of patients with LMGC who undergo systemic chemotherapy is between 7 and 14 months. However, there are currently no internationally approved standard adjuvant or palliative chemotherapy regimens for patients with advanced GC. The MAGIC trial explored the preoperative administration of epirubicin, cisplatin, and infused fluorouracil (ECF) in patients with resectable GC in the United Kingdom (18) and showed that patients receiving preoperative ECF had reduced tumour size and stage and improved overall survival (OS) compared to patients treated with surgery alone (5-year OS: 36.3% vs. 23.0%, P = 0.009). Therefore, perioperative chemotherapy (PECT) is the standard treatment strategy for patients with advanced GC in Europe and the United States. The CLASSIC trial conducted at 37 centres in South Korea, China, and Taiwan showed that postoperative capecitabine plus oxaliplatin (XELOX) regimen led to better disease-free survival (DFS) than gastrectomy alone (3-year DFS, 74.0% vs. 59.0%; P < 0.001) (19), which reinforces the efficacy of POCT in GC patients in East Asia (20, 21). SOX or XELOX regimen has been recommended as one of the standard perioperative chemotherapy regimens for advanced gastric cancer in China (22–24). Nevertheless, studies on PECT and PORT in Chinese GC patients with LM was limitied.

Multimodal chemotherapy is the preferred treatment for LMGC to improve survival. However, most studies have focused on the prognosis of patients with LMGC who also undergo hepatectomy, and there are no clinical trials or cohort studies on multimodal chemotherapy in LMGC. Therefore, predicting the prognosis of LMGC patients is difficult. We retrospectively analysed the clinicopathological and survival data of LMGC patients in the Ningxia region in China, which has a high incidence of GC. We aimed to assess the clinical features and prognostic factors of LMGC to aid in identifying the optimal chemotherapy treatment timing for these patients.

In the present study, we retrospectively collected data of 1298 patients with M1 stage GC: 546 (42.06%) in the LM group and 752 (57.94%) in the non-LM group. All patients were from the Ningxia region in northwest China, which has a high incidence of GC. The median OS of all patients was 12 months, which was consistent with data from previous reports (median OS of 3.9–18.4 months) (4–8). Our analysis further showed that the 5-year survival rate (5ySR) of the LM group was slightly lower than that of the non-LM group (9.2% vs. 10.0%), whereas the median survival time was much lower in the LM group than in the non-LM group (11 months vs. 14 months). The 5ySR of patients with LMGC in our study is much lower than that observed in previous studies examining LMGC patients from Beijing or Japan (9.2%, 12.2%, and 31.1%, respectively) (25, 26).

Several reports, including ours, have demonstrated that old age, Lauren classification, submucosal invasion, vascular involvement, lymph node metastasis, tumour differentiation, the number of LMs, and tumour size are significant factors influencing the prognosis of patients with LMGC (27–30). In the present study, we also identified POCT and palliative chemotherapy as independent prognostic factors for patients with LMGC. Compared with those receiving PECT, GC patients with LM who received POCT and palliative chemotherapy had improved survival outcomes (25.0% vs. 36.4% vs. 26.3%, P < 0.001). Palliative chemotherapy is currently the most common treatment option for LMGC, and the global standard for first-line chemotherapy is 5-fluorouracil- and cisplatin-based combinations. These chemotherapies have response rates of approximately 30-40% and a median OS of 9-11 months (31), which is lower than the median OS of 15 months in LMGC patients who received palliative chemotherapy in our study. The MAGIC trial included 503 GC patients and showed that POCT after R0 resection reduced tumour size and stage, significantly improved progression-free survival and OS, and increased the 5ySR by 13%. Therefore, POCT has been clinically considered for patients with advanced GC (32). In our study, the proportions of patients receiving PECT, POCT, and palliative chemotherapy were 3.9%, 20.4%, and 34.6%, respectively. All three modes of chemotherapy improved survival in patients with LMGC compared with those who did not receive chemotherapy, and those who received POCT had the best OS (36.4%). Moreover, with the exception of palliative chemotherapy and PECT, there was no effect on the outcomes of patients without LM, which may be due to the large proportion of patients (15.2%) with more severe peritoneal metastases, which offset the effects of chemotherapy in the non-LM group. Therefore, further prospective controlled trials are needed to investigate whether palliative chemotherapy and POCT have a survival advantage in patients with metastatic GC, especially LMGC.

Some studies have demonstrated that PECT can reduce the staging of the primary tumour, increase the possibility and efficacy of radical resection, eliminate micrometastases, and prevent or reduce tumour recurrence and metastasis in GC patients (33–35). However, the survival time of patients with LMGC did not increase with PECT in our study. The reasons for the ineffectiveness of PECT in patients with LMGC are manifold. First, gastrointestinal cancers are drained by the enterohepatic circulation, reaching the liver first, which provides a rich vascular system and sufficient nutrients for the expansive growth of tumour cells (36–38). However, the majority of chemotherapeutic drugs enter the liver through the portal vein, and liver metastases affect local drug metabolism (39). Second, LMGC patients may be less responsive to chemotherapy due to the prolonged doubling time of tumour cells, and they may be unable to tolerate adequate amounts of chemotherapy drugs (40). Last, the liver microenvironment triggers inflammation, promotes angiogenesis, and enhances permeability; therefore, PECT may not be a safe and effective intervention (41). However, these speculations are based on results derived from retrospective dataset analyses involving patients with LMGC and need to be confirmed prospectively.

The results of this study support the addition of adjuvant chemotherapy to improve OS in patients with LMGC. However, this is limited to POCT, as PECT had no survival benefit. We hope that adjuvant chemotherapy can be considered as an additional therapy for patients with LMGC in the hands of rigorous surgeons and wise oncologists. Our findings, while not prospective, provide compelling motivation to explore the potential benefits of multimodal chemotherapy in patients with LMGC in randomised clinical trials.

The limitations of this study should be considered when interpreting the results. First, all participants came from one region, and the cohort may not be representative of all GC patients in China. Second, the study reviewed eight years of clinical data on patients with LMGC who received any combination of chemotherapy (administered orally, intravenously, or intraperitoneally). Patients were given SOX or XELOX perioperative chemotherapy regiments, which was divided into 2– 4 cycles, the evaluation results may be biased. Therefore, further investigation in multiple populations is required. Third, the results are dependent on the quality of data collection. In addition, there were only 57 LMGC patients in the PECT group, and the conclusion that PECT has no prognostic significance in patients with LMGC needs to be verified in larger studies.

In conclusion, this study collected 1298 patients with metastatic GC from a high GC incidence area in China, examined a variety of clinical characteristics to construct a Cox proportional hazards model, and systematically evaluated the prognostic factors that may affect patients with LMGC. Our study revealed that LMGC patients had a worse prognosis than non- LMGC. Number of metastatic sites more than 1, liver and other metastatic sites, no CT treatment and HER2-negative were all unfavorable factors to the survival of LMGC. LMGC patients who received palliative chemotherapy and POCT had significantly better survival than those who received PECT. These prognostic factors can help clinicians and patients quantify the benefits of adjuvant therapy for LMGC and make individualised treatment recommendations and decisions.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by Ningxia Medical University. The patients/participants provided their written informed consent to participate in this study.

All the authors were involved in the design of the study or in the explanation of the data. XL, ZC and YZ together collected the data of the patients, and sorted out, analyzed and evaluated the quality of the data, which was checked by GW and WY. ZC and YZ contributed to the analysis and write-up of findings. XL wrote the first draft of the manuscript. CZ and HZ contributed to the methodology. HN, HQ and XJ contributed to proofread the article. All authors contributed to the article and approved the submitted version.