Adenocarcinoma of esophagogastric junction (AEG) refers to adenocarcinoma where the tumor center is located within 5 centimeters above or below the esophagogastric junction and crosses or contacts this line. In recent years, the epidemiological trends of AEG have shown notable changes, primarily attributed to the accelerated aging of the global population, diversified lifestyles, and complex environmental factors. According to an analysis of global gastric and esophageal cancer registry data from 2018, East Asia accounted for as much as 67.1% of new cases of AEG globally, showing a significant increase compared to data from 2012 (1, 2). The incidence of AEG in China has also shown an upward trend, consistent with the trends in Western countries, and the proportion of early cases was always less than 20% (3–5). Most patients are diagnosed at a locally advanced stage, and surgical resection remains the primary treatment for such cases. However, after surgery alone, patients often face high risks of recurrence and metastasis, leading to poor outcomes. Neoadjuvant chemotherapy (nCT) or neoadjuvant chemoradiotherapy (nCRT) can reduce tumor size, increase R0 resection rates, and extend recurrence-free survival (RFS) and overall survival (OS). These benefits have been confirmed by numerous large-scale clinical studies. nCT or nCRT combined with surgical resection is gradually becoming the standard treatment model for locally advanced AEG.

In clinical practice, the combination of programmed cell death ligand 1 (PD-L1) inhibitors and chemotherapy has achieved significant results in the treatment and long-term prognosis of patients with advanced upper gastrointestinal tumors, including AEG (6–8). This approach provides a new first-line treatment option for patients with advanced AEG. The oncological efficacy advantages obtained in advanced upper gastrointestinal tumors have also sparked enthusiasm for exploring new adjuvant immunotherapy regimens. Some prospective phase II and phase III clinical trials have confirmed that preoperative nCI shows notable benefits in terms of safety and improved pathological response rates in gastric or esophageal cancer (9–12).

Lately, the impressive findings from the KEYNOTE-062 and CheckMate-649 clinical trials (13, 14) have confirmed the therapeutic efficacy of immunotherapy in combination with chemotherapy for unresectable advanced gastric/gastroesophageal junction(G/GEJ) cancer.These positive outcomes provide a solid rationale for exploring the application of this treatment strategy in the neoadjuvant therapy setting for patients with advanced G/GEJ cancer. While,Siewert type II AEG is demarcated by a virtual anatomical boundary within the esophagogastric junction region, and its unique anatomical location particularity entails a host of complexities and challenges,with prognosis influenced by tumor stage, nodal status, and resection margins (15). There is considerable controversy regarding its treatment strategies, including the pattern of lymph node metastasis, surgical approach, and extent of resection. Besides, the efficacy of nCI in Siewert type II AEG has not yet been established and lack of large-sample data support. This study aims to verify the efficacy and safety of nCI in real-world settings for locally advanced resectable Siewert type II AEG, and to further analyze the factors related to efficacy and safety.

At present, the standard neoadjuvant treatment for locally advanced Siewert type II AEG remains based on chemotherapy or chemoradiotherapy with fluorouracil combined with platinum agents. The MAGIC study (16)demonstrated that perioperative ECF chemotherapy improved overall survival compared with surgery alone, with a 5-year OS of 36% versus 23%. The EORTC 40954 trial (17) confirmed that neoadjuvant FLP chemotherapy improved R0 resection rates (81.9% vs. 66.7%) but did not show a survival benefit. More recently, the FLOT4-AIO trial (18)established the FLOT regimen as a standard perioperative option, with a pCR rate of 16% and an R0 resection rate of 85%。

Specifically for Siewert II/III AEG, retrospective analyses have shown even lower pathological responses with nCT alone, with pCR rates around 3% and R0 resection rates of 82%, leading to inferior survival compared with nCRT (3-year OS 58.0% vs. 79.2%) (17, 19). Although nCRT can improve tumor regression, it is often associated with severe toxicities that compromise tolerance and compliance. As a result, many patients fail to complete the planned regimen, potentially missing the opportunity for curative surgery. Therefore, more effective and tolerable treatment strategies are urgently needed to enhance efficacy and prognosis in locally advanced Siewert type II AEG.

In recent years, immunotherapy has brought new breakthroughs, several studies such as ORIENT-16, ATTRACTION-4, CheckMate-649 (8, 13, 20) have indicated that the combination of immunotherapy with chemotherapy can improve the efficacy and prognosis of advanced gastric cancer, making it a new option for the treatment of locally advanced AEG. Several small-sample phase II clinical trials (12, 21), such as NEOSUMMIT-01 and NEOSUMMIT-03, have demonstrated that neoadjuvant chemotherapy combined with immunotherapy shows promising application prospects in terms of pCR rate, MPR rate, and safety for locally advanced gastric cancer or esophagogastric junction cancer. However, most of these studies involved small sample sizes and did not exclusively enroll Siewert type II AEG patients,the value of nCI in locally advanced Siewert type II AEG remains unclear. Therefore, we conducted this study to evaluate the feasibility and safety of a new neoadjuvant treatment regimen combining chemotherapy with ICIs for patients with resectable locally advanced Siewert type II AEG. We observed a pCR rate of 18.8%, an MPR rate of 35.6%, and an R0 resection rate of 94.1%. Compared with historical nCT trials, these results suggest at least comparable, and potentially superior, pathological outcomes, although cross-trial comparisons should be interpreted with caution.

This study demonstrates that the neoadjuvant treatment regimen combining fluorouracil/tegafur with platinum-based chemotherapy and ICIs shows significant efficacy for resectable locally advanced Siewert type II AEG. Overall, the pCR rate was 18.8%, the MPR rate was 35.6%, and the pathological downstaging rate was 73.3%. Additionally, among clinical T stages, patients with cT3 stage showed better outcomes compared to those with cT4 stage in terms of pCR, MPR, and postoperative pathological downstaging (pCR 27.9% vs 12.1%, p=0.044; MPR 48.8% vs 25.9%, p=0.017; postoperative pathological downstaging rate 83.7% vs 65.5%, p=0.041). At present, the assessment of tumor regression response is of notable importance for determining treatment efficacy, formulating subsequent treatment plans, and predicting patient prognosis (22). It also serves as a good predictor for recurrence, metastasis, and survival following neoadjuvant therapy for tumors (23–25). The pCR and MPR rates observed in this study (18.8% and 35.6%, respectively) were higher than those observed in some previous clinical studies of nCT and neoadjuvant monotherapy immunotherapy, which ranged from 4.0% to 6.3% (26–29) and 16.1% to 30% (18, 29). This suggests that the nCI may be more effective than chemotherapy or immunotherapy alone. This may be attributed to the regulatory effect of chemotherapy drugs on the immune status of tumor microenvironment, thereby promoting the release of tumor antigens (30–32). The synergistic effect of immunotherapy and chemotherapy has likely contributed to the higher pCR rates and MPR rates.

With a median follow-up of 19.2 months, the median RFS and OS had not yet been reached, underscoring the relatively immature survival data at this stage. Accordingly, the present results should be regarded as an interim analysis. Patients in the MPR group experienced significantly better RFS compared with the non-MPR group (p=0.037). In contrast, although the pCR group showed a numerical trend toward improved RFS (p=0.088), this difference did not reach statistical significance. The cumulative incidence of tumor-related death further highlighted this pattern: in the non-MPR group, it was 5.4% (95% CI: 0.0–11.5%) at 1 year and 26.8% (95% CI: 11.1–42.6%) at 2 years, whereas no tumor-related deaths occurred in the MPR group at either timepoint (Gray’s test, p=0.008). However, no significant difference was observed between the pCR and non-pCR groups. Several factors may account for this discrepancy. First, the relatively small sample size reduces the statistical power to detect survival differences, particularly when stratifying by pathological response. Second, the short follow-up period, with median RFS not yet reached, results in immature survival data. Third, non-cancer-related deaths and the impact of postoperative adjuvant therapy may have diluted the prognostic effect of pCR. Nevertheless, the observed association between pathological response and improved outcomes is consistent with previous clinical studies (25, 33). Taken together, these findings should be interpreted with caution, and the prognostic value of pCR in this setting requires validation in larger, prospective trials with longer follow-up.

There is no unified standard for the number of neoadjuvant treatment cycles for resectable locally advanced Siewert type II AEG, leading to much controversy. Most Japanese studies adopt a two-cycle regimen, whereas studies such as FLOT4, MAGIC, and FFCD (16, 34, 35)opt for 3–4 cycles and consider it beneficial for patients. Our research results show that patients who completed 3–4 cycles of neoadjuvant therapy had a better pCR rate compared to those who completed 1–2 cycles of neoadjuvant therapy (p=0.015). However, there was no significant difference in the MPR rate and postoperative pathological downstaging rate. This finding may be due to the fact that two cycles of neoadjuvant therapy can achieve the goals of downstaging and downgrading, but if a higher pCR rate is desired, the number of neoadjuvant cycles needs to be increased. The selection of the number of cycles for neoadjuvant therapy in resectable locally advanced Siewert type II AEG should comprehensively consider various factors such as chemotherapy regimen, efficacy, adverse reactions, and drug efficacy. These factors all determine the choice of the number of nCI cycles, which in turn affects the patient’s treatment outcomes and clinical results.

Regarding the safety of preoperative treatment, the incidence of TRAEs in this study was 35.69%, with the most common hematological toxicity being leukopenia (15.84%),which less than that of neoadjuvant ECF/ECX and FLOT formula reported in the FLOT4 study (34). and the most common non-hematological toxicity being nausea and vomiting (16.83%). Most TRAEs were classified as grade I-II, and no treatment-related deaths occurred in this study. Compared with nCT related studies (16, 17, 27, 34), there was no additional risk of treatment. However,7 patients occurred irAEs, 3patients(2.97%) experienced grade 3–4 irAEs, including 1 case of Immune myocarditis (0.99%), 1 cases of Immune myositis (0.99%), and 1 case of Immunologically associated thyroid dysfunction (0.99%). This study highlights that myocarditis, though rare, can occur with a delayed onset (3 weeks after treatment initiation) and requires prompt recognition (via troponin/CK-MB monitoring) and multidisciplinary management (glucocorticoid therapy + cardiology collaboration). Consistent with guidelines from the American Society of Clinical Oncology (ASCO) for irAE management, early intervention led to favorable outcomes without long-term cardiac sequelae, providing practical experience for clinical practice.Severe irAEs, though rare (2.97%), require vigilant monitoring—particularly myocarditis, which, while occurring in only 1 patient here, carries life-threatening risks.Overall, these adverse events are considered safe and manageable, and no treatment-related deaths have occurred. ICIs treatment for AEG has resulted in significant therapeutic improvements, yet it is also accompanied by a series of unique adverse events. This necessitates that clinicians possess a high degree of vigilance and expertise, as meticulous monitoring and timely intervention. Most adverse events can be properly managed, thereby ensuring patient safety. In the future, with the deepening of research into the toxicity mechanisms related to ICIs, we can expect to develop more precise predictive tools and individualized treatment plans, to further optimize the application of ICIs in the treatment of AEG, and to enhance the overall treatment outcomes and quality of life for patients.

In our study, most patients (90.1%) underwent “Upper abdominal + Left Thoracotomy”, which is the prevailing practice at our center. Data from previous concurrent studies (36) at our center showed that the lymph node metastasis rate of Siewert type II AEG is highest in the upper perigastric + supra-pancreatic region (61.6%), followed by the lower perigastric + hepatoduodenal region (17.1%) and the lower mediastinum (5.4%), with no metastasis observed in the middle and upper mediastinum. This multi-regional lymph node metastasis pattern necessitates a surgical approach capable of comprehensively addressing both thoracic and abdominal lymph nodes, and the “Upper abdominal + Left Thoracotomy” approach precisely meets this requirement. Compared with the traditional Sweet procedure or the upper abdominal approach, this surgical method does not increase the incidence of postoperative complications. Nevertheless, a small proportion of patients in this study cohort underwent the upper abdominal approach, which was determined based on individualized selection according to the patients’ physical conditions and tumor characteristics.Currently, the optimal surgical approach for Siewert type II AEG remains debated.We emphasize that our findings reflect the outcomes of our institutional approach, and further prospective studies comparing surgical techniques (e.g., randomized trials of open vs. minimally invasive resection for Siewert type II AEG) are needed to define the optimal approach.

In addition, the incidence of surgical-related complications was 36.6%, with the most common postoperative complication being lung infection (28.7%), followed by deep vein thrombosis in the lower limbs (4.9%). The postoperative complications were relatively manageable, and no perioperative mortality events occurred. The average operative time was 266 minutes, with an average intraoperative blood loss of approximately 150ml, a median number of lymph node dissections of 28, and an average postoperative hospital stay of 9 days. There were no intraoperative complications, consistent with previous studies (27, 37) reporting on nCT or nCRT combined with surgical treatment. The results confirm the safety and feasibility of nCI in resectable locally advanced AEG patients, without increasing the difficulty and risk associated with surgery.

Our study has several inherent limitations. First, it was a single-center retrospective analysis without a contemporaneous nCT control group, which may introduce selection bias and limits the strength of causal inference. Second, the follow-up duration was relatively short, the survival endpoints remain immature, and the observed advantages of nCI should be regarded as preliminary; Longer follow-up is essential to fully evaluate its impact on long-term survival and recurrence. Third, biomarker analyses were limited. Routine testing for PD-L1 expression, MSI, TMB, and dMMR was not performed in all patients due to the retrospective nature of this study. Ongoing prospective studies at our center are expected to provide more comprehensive biomarker data and long-term clinical outcomes, thereby strengthening the confirmation of our preliminary results.

Neoadjuvant chemotherapy combined with immunotherapy show a promising efficacy in the treatment of resectable locally advanced Siewert type II AEG, with high rates of pCR and MPR, as well as good tolerance and safety. Moreover,3–4 cycles of nCI may provide better therapeutic efficacy than 1–2 cycles. Additionally, this treatment regimen has potential long-term benefits in reducing recurrence and metastasis rates and prolonging survival.Validation in larger prospective studies with extended follow-up is required to determine the long-term survival impact of this regimen.