In 2022, gastric cancer (GC) became the fifth most common cancer and the fifth leading cause of cancer-related death worldwide (1). Although the incidence and mortality of GC have declined over several decades in most parts of the world, the disease burden of GC remains high in Asia. In China, GC is the fifth most frequently diagnosed cancer and the third leading cause of cancer-related death (2). GC typically occurs in middle-aged and elderly patients, especially those aged ≥65 years, but can occur in young patients, although it is largely ignored in the clinic. Recent findings indicate an increasing trend in GC incidence rates among young patients. At present, there is no clear age limit for young patients with GC. In some studies, young GC is defined as GC before age 40, while in others, the definition has included all patients diagnosed before age 45 (3–6). Whether the cut-off is 40 or 45 years, the incidence of young GC patients has increased annually (7, 8). The occurrence of GC in young patients is less affected by environmental factors and more strongly related to heredity (9, 10). Compared with elderly patients, young patients commonly have more aggressive tumours, more advanced disease at presentation, and worse prognoses (11). Meanwhile, due to the lack of obvious early symptoms and delayed diagnosis, young patients often present at an advanced stage when seeking medical care, resulting in the loss of opportunities for radical surgery (12). Therefore, actively exploring pathological molecular characteristics and advancing precision therapy for this condition are of critical significance.

In recent years, immune checkpoint inhibitors (ICIs) have achieved remarkable success across various cancer types, offering a promising advancement in the treatment of GC. Supported by global multicentre phase III clinical trials, programmed death 1 (PD-1) inhibitors combined with chemotherapy have been approved as the standard first-line therapeutic regimen for advanced GC (13–16). However, current treatment strategies for advanced GC remain undifferentiated by age, and no consensus clinical guidelines specifically address the management of young patients in international practice. A recent study demonstrated that immune responses differ with sex and age, with female and young patients having lower response rates to immunotherapy (17). In another multi-institutional retrospective study including 538 patients with advanced melanoma, the association between age and the response to anti-PD1 therapy was evaluated. The study showed that patients younger than 62 years were less likely to benefit from pembrolizumab treatment, with a disease control rate of 52% compared to 63% in older patients. The investigators also noted that the likelihood of response increased with age and that the odds ratios of progressing on pembrolizumab decreased 13% with every decade of patient age (18). Therefore, the efficacy of immunotherapy in young patients still needs to be clarified.

Our previous study demonstrated that young gastrointestinal patients (aged <45 years) receiving PD-1 inhibitor therapy exhibited significantly suboptimal treatment responses and inferior survival outcomes compared to elderly cohorts (19). In this study, we retrospectively analysed the clinical data of young GC patients aged ≤45 years, summarized their clinicopathological characteristics, and compared the efficacy of different first-line treatment schemes, especially PD-1 inhibitors combined with dual chemotherapy, to further verify the efficacy of immunotherapy in young patients and provide data supporting the realization of precision treatment for young GC patients.

The incidence of GC in young individuals has increased steadily in recent decades (7, 8). Previous research has revealed that the detailed clinicopathologic features and prognostic factors of young patients with GC are still debated. Therefore, we conducted the current study to investigate a young group of patient under the age of 45 years. In our analyses, female patients constituted a higher proportion of the entire group, with a male-to-female ratio of approximately 1:1.56. Previous reports have shown that this incidence of GC is higher among males and patients of advanced age, but among younger patients, the incidence is higher among females (20–22). Our study supports this finding of a higher proportion of female patients among young patients. The reason for the female predominance among young patients remains unclear. Several studies have suggested that sex hormones, especially oestrogen, may play an important role in the development of GC in young people (23). Zhong et al. showed that oestrogen-related receptors are highly expressed in GC tissues, where they can promote the migration and invasion of cancer cells (24). Harrison et al. found that oestradiol is trophic to GC cell lines in vitro and suggested that positivity for oestradiol receptor D5 expression is an independent negative prognostic factor for patients with GC (25). In addition, sex hormones and sex chromosome-related genes are also the main factors driving these differences in immunity (26, 27). The molecular mechanisms underlying sex-based differences in nonresponsiveness to ICIs, such as oestrogen-mediated recruitment of myeloid-derived suppressive cells (MDSCs) and Tregs to the TME, which are known to be involved in resistance to ICIs, have recently been characterized in mouse models (28). Interestingly, our study found that 41.3% of all female patients with a history of childbirth underwent caesarean section, while the rate of caesarean section in China is only 34.9% (29). The relatively high proportion of patients with GC who underwent caesarean section suggests that caesarean section may be related to the occurrence of GC. However, there are currently few studies on the relationship between the medical history of female patients and GC, and further research is necessary to explore this further. In summary, GC can occur in young female patients, especially those who undergo caesarean section, and symptoms suggestive of GC should be investigated aggressively.

Currently, the occurrence and progression of GC are multifactor and multistep processes that are the result of interactions between the environment and heredity. However, the incidence of GC among young patients is less affected by environmental factors and more related to heredity (9, 10). Familial clustering was found in 11.4% of GC patients and was more notable in young patients (30). In our study, 14.2% of GC patients had a family history of tumours, consistent with other reports (6, 31). The most common hereditary GC is hereditary diffuse gastric cancer (HDGC), which is an autosomal dominant genetic syndrome often caused by E-cadherin gene (CDH1) mutation. Decreased or dysfunctional e-cadherin can lead to decreased adhesion between cells, increasing tumour aggressiveness (32, 33). Due to the limited research scope, we did not sequence the exons of tumour tissues. Nonetheless, it is important to carry out regular endoscopic screening for young patients who have a family history of tumours or genetic mutations who are at risk of GC.

In addition, recent studies have shown that young GC patients have a different molecular expression profile than old patients, including a lower frequency of HER2 overexpression and mismatch repair (MMR) deficiency (34, 35). In our cohort, HER2 status was examined in 172 patients, with 12.2% showing positive expression. According to the HER-EAGLE study, only 9.2% HER2 positivity was detected in patients before age 55 (36). This difference from our study may be explained by the study era and geographic variations. Recent studies have shown that HER2 itself regulates the recruitment of tumour-infiltrating immune cells by inducing the expression of chemokine ligand 2 (CCL2) and PD-1 ligands, thereby enhancing the efficacy of immunotherapy (37). The KEYNOTE-811 study showed that pembrolizumab combined with trastuzumab and chemotherapy increased the ORR of patients with HER2-positive advanced GC to 74.4%, which was significantly higher than that of the control group. Based on this study, immunotherapy combined with anti-HER2 therapy plays an important role in the first-line treatment of advanced GC (16). Microsatellite status was detected in 96 patients, with only 1 of them having dMMR/MSI-H in our survey. The low incidence of MMR deficiency in young patients with GC is probably related to the high proportion of diffuse tumours, wherein MSI is less common, and to the fact that genomically stable tumours are usually diagnosed at an earlier age (38). Patients with MMR deficiency tend to generate more neoantigens and are more immunogenic, thus responding better to ICIs (39). At present, MSI status is the most accurate molecular marker for predicting the efficacy of immunotherapy. In addition, PD-L1 expression was evaluated in 93 patients in our study, 43.0% of whom exhibited positive PD-L1 expression. PD-L1 positivity was defined as a combined positive score (CPS) ≥1. Rha SY et al. included a large cohort of 574 GC patients and demonstrated that 67.4% of all patients had a CPS ≥1 (40). Another analysis showed a positive correlation between PD-L1 expression and the CPS and ICIs response in patients with GC (41). PD-L1 expression has rarely been reported in young GC patients, so these results may provide a reference for further exploration. In conclusion, young GC patients may not be good candidates for immunotherapy, so novel therapies involving different approaches are needed.

Recently, based on the phase III clinical trials, PD-1 inhibitors combined with dual chemotherapy have become the standard first-line treatment for advanced GC patients (13–16). However, therapeutic options for GC are not stratified by age worldwide, and young GC patients may not be good candidates for immunotherapy. In a recent study evaluating the relationship between age and immunotherapy in patients with lung adenocarcinoma, the tumour mutation burden (TMB) of patients in the older group (>50 years) was significantly higher than that of patients in the younger group (≤50 years), and the OS of patients receiving immunotherapy was longer in the older group (42). Specific genetic mutations also influence responses to ICIs. Integrating genomic profiles from TCGA and clinical immunotherapy data from MSKCC, Guan R et al. identified age-related mutational disparities: young patient cohorts exhibited higher frequencies of IDH1, BRAF, and ATRX mutations, whereas TP53, TTN, MUC16, and LRP1B mutations were more prevalent in older cohorts (43). Notably, BRAF mutations appeared to confer inferior responses to ICIs, while patients with LRP1B mutations exhibited significantly higher TMB and improved clinical outcomes compared to wild-type counterparts (44, 45). We also previously reported that PD-1 inhibitors therapy has lower efficacy against gastrointestinal cancer in younger patients (aged <45 years) (18). In this study, we focused on young unresectable GC patients, compared the efficacy of different first-line treatment regimens, and found that PD-1 inhibitors plus dual chemotherapy were not superior to dual chemotherapy alone. The low response rate to anti-PD-1 therapy in young GC patients is not only related to the molecular expression profile of young GC patients themselves but also may be due to the higher incidence of young GC in female patients. A recent study revealed that young and female patients accumulate driver mutations in their tumours that are less readily presented by their major histocompatibility complex (MHC) molecules (17), further suggesting that these effects are strong and complementary. Together, current knowledge provides a rationale for the paradigm that immune selection exerts its Toll differently concerning age and sex, with a strong immunoediting effect being observed in young and female patients. In addition, this study found that 49.3% of young GC patients had peritoneal metastasis at the first visit, which is a negative clinical indicator of immunotherapeutic efficacy. In a multicentre biomarker cohort study on the efficacy of nivolumab treatment for GC, the results showed that patients with nonperitoneal metastases had a higher response rate to immunotherapy than those with peritoneal metastases (46). Previous studies have pioneered the integration of machine learning and single-cell sequencing analysis to systematically investigate the mechanisms underlying immunotherapy efficacy in lung adenocarcinoma (47–49). Our future studies will employ scRNA-seq and spatial transcriptomics to map age-stratified immune landscapes. Coupled with deep learning models (50), this will identify novel targets such as exhausted CD8⁺T cell subsets or immunosuppressive CD8⁺ Treg populations, which may underlie the attenuated immunotherapy efficacy in young patients.

However, young patients are in relatively good physical condition, have few complications, have high tumour malignancy, experience rapid disease progression, and need and tolerate more aggressive treatment options. In the present study, we compared the efficacy and prognostic impact of different first-line treatment regimens in young unresectable GC patients. Triple regimens provided a significant DCR and PFS advantage compared with dual chemotherapy and PD-1 inhibitors plus dual chemotherapy. Although baseline imbalances in patient characteristics, temporal variations, and genetic heterogeneity may compromise the generalizability of the findings, this study provided compelling evidence supporting that these young GC patients may benefit from a stronger chemotherapy regimen. Next, we performed a more detailed survival analysis based on different treatment regimens and found that intraperitoneal infusion of paclitaxel followed by intravenous paclitaxel combined with S-1 and apatinib oral therapy had better PFS and OS than other regimens, which may be related to the higher proportion of patients with peritoneal metastasis in this study. For patients with peritoneal metastasis, due to the existence of a plasma-peritoneal barrier, it is difficult for chemotherapeutic drugs with large molecules to penetrate the barrier to act on metastases in the abdominal cavity, so systemic chemotherapy has little effect on patients with peritoneal metastasis. Intraperitoneal perfusion chemotherapy provides a new treatment option for patients with peritoneal metastasis. Meanwhile, young GC patients frequently present as a poorly differentiated or undifferentiated adenocarcinoma, characterized by heightened tumour cell proliferative activity and a larger proportion of cells residing in active phases of the cell cycle. This aggressive proliferative phenotype may render these tumours more sensitive to paclitaxel, a taxane-class chemotherapeutic agent that targets rapidly dividing cells by stabilizing microtubules and disrupting mitotic spindle dynamics, thereby inducing cell cycle arrest and apoptosis (6). PHOENIX series studies in Japan have confirmed that intraperitoneal infusion of paclitaxel followed by intravenous paclitaxel combined with S-1 oral therapy is a safe and feasible treatment method for GC with peritoneal metastasis (51). In addition, apatinib, as a highly selective VEGFR-2 inhibitor, exhibits fewer off-target effects compared to multi-targeted TKIs, enhancing its therapeutic index in GC (52). Therefore, intraperitoneal infusion of paclitaxel followed by intravenous oral paclitaxel combined with S-1 and apatinib may be recommended as a preferred option for young patients. However, higher quality trials, better patient selection, and multicentre randomized controlled trials are needed to support this conclusion. In summary, there is no doubt that young GC patients have fewer complications, better PS, and adaptability to aggressive treatment. Our results may support an argument that is worth verifying in future research that young GC patients should receive more aggressive treatment.

Our study has several limitations. First, this study was retrospective in nature. Second, the relatively small sample size limited the reliability of the conclusions. Incomplete data on molecular markers such as HER2, MMR status and PD-L1 expression, particularly the unexplored association between PD-L1 expression levels and the therapeutic efficacy of immunotherapy in young GC patients, may introduce bias into the study outcomes. Nevertheless, this study revealed critical clinical characteristics of young GC patients, including a female predominance, lower rates of HER-2 positivity, reduced prevalence of PD-L1 expression, diminished proportions of MSI-H/dMMR, and a higher incidence of peritoneal metastasis. Furthermore, our exploration provided real-world data for the precision first-line treatment of young patients. Prospective validation of age-stratified treatment protocols, particularly in immunotherapy trials, is urgently warranted to translate these findings into clinical practice.

This study revealed important clinicopathological features of young unresectable GC patients, which may lead to a poor response to ICIs treatment in this population. In terms of first-line treatment in young patients, PD-1 inhibitors plus dual chemotherapy were not superior to dual chemotherapy alone, but triple regimens, especially intraperitoneal infusion of paclitaxel followed by intravenous paclitaxel combined with S-1 and apatinib oral therapy, were superior to other regimens. To our knowledge, this is the first study to explore first-line treatment specific to the population of young GC patients. Prospective new studies are needed to provide more information on this subject.