Gastric cancer is a common malignant tumor that seriously threatens human health, and its incidence and mortality remain high in China (1). With the aggravation of aging population, the number of elderly patients with gastric cancer increases. The decline of body function of this group, often accompanied by a variety of chronic diseases, makes the postoperative clinical outcome complex and the prognosis is different. Therefore, accurate prediction of postoperative clinical outcomes in elderly patients with gastric cancer is essential for formulating personalized treatment options and improving the prognosis (2). The neutrophil-to-lymphocyte ratio (NLR) in peripheral blood, as a key indicator reflecting systemic inflammatory and immune status, has emerged as a crucial biomarker for cancer prognosis (3). Elevated NLR levels demonstrate enhanced systemic inflammation and immune suppression, thereby mirroring tumor cell proliferation, invasion, and metastatic potential. For elderly gastric cancer patients with inherently compromised immune function, NLR serves as a sensitive marker for evaluating postoperative anti-tumor immunity and recovery capacity (4). These properties justified its inclusion as a study parameter. However, NLR alone cannot comprehensively predict postoperative outcomes due to the multifactorial nature of clinical results influenced by additional variables. Carcinoembryonic antigen (CEA) is a common tumor marker and has great significance in the diagnosis and treatment of gastric cancer (5). When gastric cancer occurs, tumor cells secrete CEA, resulting in an increase in its level in the blood. High CEA levels are often associated with high tumor invasiveness and poor prognosis (6). In addition, factors such as age, tumor TNM stage, lymph node metastasis, tumor differentiation degree and comorbidities can also significantly affect the prognosis of patients with gastric cancer. Elderly patients have poor tolerance to surgery and treatment. Late tumor staging, multiple lymph node metastases, low degree of differentiation and the combination of chronic diseases will increase the risk of surgery and affect rehabilitation. At present, there are few studies on the prediction model of clinical outcomes after surgery for gastric cancer in the elderly constructed by combining NLR and CEA with the above various factors. This study aims to fill this gap, build a multi-factor Nomogram prediction model, provide an accurate prediction tool for clinical practice, guide the treatment and management of elderly patients with gastric cancer, and improve the survival rate and quality of life of patients.

In this study, we constructed and validated a Nomogram prediction model for clinical outcome of elderly patients with gastric cancer after surgery based on multiple factors, including NLR, BMI, tumor size, lymph node metastasis, CEA and age. This result is of great significance for prognostic evaluation and clinical management of elderly patients with gastric cancer. The developed nomogram provides clinicians with a practical tool for individualized risk assessment. To operationalize this model: (1) For each predictor (x1-x6), locate the patient’s value on the corresponding axis and draw a vertical line to the ‘Points’ axis to determine the partial score. (2) Sum all partial scores to obtain the Total Points. (3) Locate the Total Points on the bottom axis and draw a vertical line to the ‘Disease Risk’ axis to read the predicted probability. For illustration, consider a patient with the following profile: NLR (x1)=3 (50 points), BMI (x2)=23 kg/m² (40 points), Tumor size (x3)=1.8 cm (20 points), Lymph node metastasis (x4)=0 (25points), CEA (x5)=5.7 ng/mL (45 points) and Age (x6)=74 years (40 points). The cumulative score of 220 points corresponds to an estimated 50% probability of poor clinical outcomes. This threshold-based stratification enables clinicians to identify high-risk patients who may benefit from enhanced surveillance or early intervention, and optimize resource allocation while maintaining prognostic accuracy.

Based on the nomogram’s total point distribution (0–240 points), we established three clinically actionable risk categories: low-risk (≤155 points, disease risk <30%), intermediate-risk (155–175 points, 30–60% risk), and high-risk (>175 points, >60% risk). Clinically, this stratification guides management: low-risk (standard follow-up), intermediate-risk (enhanced surveillance), and high-risk (adjuvant therapy consideration). The thresholds balance simplicity with individualized accuracy, aligning with the nomogram’s predictive power.

The nomogram’s visual design and manual scoring system enable rapid risk assessment (<5 minutes) during bedside evaluations, supporting immediate clinical decisions (e.g., adjuvant therapy planning or intensified follow-up for high-risk cases). For streamlined workflows, electronic health record (EHR) systems allows automatic extraction of patient data (age, NLR, BMI, tumor size, lymph node metastasis status, CEA) and generates real-time risk predictions, ensuring scalability across healthcare settings. Crucially, the model is exclusively applicable to elderly gastric cancer patients meeting this study’s inclusion criteria: those undergoing radical surgery without other malignancies or severe organ dysfunction (e.g., severe heart/liver/kidney failure), with complete preoperative NLR and tumor-specific data (including verified tumor size, lymph node metastasis status, and CEA levels). It is not recommended for patients with incomplete clinical data (e.g., missing preoperative NLR) or those receiving non-surgical treatments (e.g., palliative care for advanced unresectable tumors), as predictive performance remains unvalidated in these populations.

NLR, as an indicator reflecting the inflammation and immune state of the body, has been shown to have a significant impact on the clinical outcome of postoperative gastric cancer in the elderly in this study. The occurrence and development of tumors are closely related to the imbalance of inflammatory microenvironment and immune function (8). The results of a previous meta-analysis provided strong or highly suggestive evidence supporting the association between NLR and cancer prognosis (8). Previous study has suggested that NLR might be particularly pertinent to the prognosis of GC patients. In conclusion, the inflammatory markers NLR, PLR, and LMR serve as effective biomarkers for prognostic assessment in GC patients (9). The association between elevated NLR and poor postoperative prognosis may be attributed to its role in shaping an immunosuppressive tumor microenvironment (10). The increased NLR implies enhanced inflammatory response and immunosuppression of the body, which may create favorable conditions for the proliferation, invasion and metastasis of tumor cells (11). In elderly patients with gastric cancer, the immune function is inherently relatively weak due to body function decline. The increased NLR further aggravates the immune disorder, which is not conducive to postoperative recovery and increases the risk of adverse clinical outcomes (12).

BMI is also an important factor affecting the clinical outcome of postoperative gastric cancer in the elderly. Obesity (higher BMI) has been associated with poor prognosis in many cancers, as well as in elderly patients with gastric cancer (13). On the one hand, obese patients have excessive adipose tissue, which can secrete a variety of cytokines and hormones, such as leptin and adiponectin, which regulate the growth, proliferation and invasion of tumor cells. Leptin can activate signaling pathways in tumor cells, promote cell proliferation and resist apoptosis. Abnormal adiponectin levels may affect the metabolism and immune escape of tumor cells. On the other hand, obesity increases the difficulty of surgery and the risk of postoperative complications, such as wound infection and pulmonary infection, and thus affects the patient’s rehabilitation process, leading to adverse clinical outcomes (14).

The tumor size directly reflects the tumor load degree. Larger tumors tend to mean a longer course of disease, greater invasiveness, greater susceptibility to invasion of surrounding tissues and organs, and greater likelihood of distant metastasis (15). For elderly patients with gastric cancer, the body has a poor tolerance to tumors. The larger the tumor, the more serious the consumption of the body. The greater the difficulty and risk of surgical resection. Even with surgical treatment, the potential for residual tumor cells is relatively high, which undoubtedly increases the risk of postoperative recurrence and adverse clinical outcomes (16). In clinical practice, for elderly patients with large tumors, the feasibility of surgery should be carefully evaluated and a comprehensive treatment plan should be formulated to improve the prognosis.

Lymph node metastasis is an important pathway for tumor metastasis and also one of the key factors affecting the prognosis of patients with gastric cancer (17). Once lymph node metastasis occurs, the tumor cells have crossed the barrier of local tissue and entered the lymphatic circulation system, which indicates a higher risk of recurrence and a worse prognosis. Elderly patients with gastric cancer have a weak resistance to tumor metastasis due to decreased body function, and it is more difficult to control the progression of the disease through autoimmune and therapeutic means after lymph node metastasis. In addition, the number and location of lymph node metastasis are also closely related to the prognosis. the more metastatic lymph nodes and the farther the location, the shorter the survival time of patients and the higher the incidence of adverse clinical outcomes (18). The lymph node metastasis should be fully considered in the development of treatment plan, and postoperative adjuvant treatment should be strengthened to reduce the risk of recurrence.

As a common tumor marker, CEA has important significance in the diagnosis and treatment of gastric cancer (19). When gastric cancer occurs, tumor cells secrete CEA, resulting in an increase in its level in the blood. High CEA levels are often associated with high tumor invasiveness and poor prognosis. In this study, CEA was also an independent risk factor for poor clinical outcomes in elderly patients after gastric cancer surgery (20). CEA can not only serve as a marker for tumor presence but also reflect tumor biological behavior. Elevated CEA levels may indicate enhanced proliferative, invasive, and metastatic capacities of tumor cells, and may also reflect CEA-mediated immunosuppression in the tumor microenvironment—such as impaired T-cell cytotoxicity or increased regulatory T-cell infiltration, which weakens the ability of body to clear residual tumor cells. Monitoring changes in CEA levels can help to evaluate the efficacy of treatment and predict the risk of recurrence, which is of great value for dynamic management of elderly patients with gastric cancer.

Age is a factor that cannot be ignored in elderly patients with gastric cancer. With the growth of age, the functions of various organs of the body of elderly patients gradually decline, and their tolerance to surgery and treatment is significantly reduced (21). Elderly patients are often accompanied by a variety of chronic diseases, such as hypertension, diabetes, coronary heart disease, and so on. These complications will increase the risk of surgery and the incidence of postoperative complications. At the same time, the elderly will also lead to decreased immune function of the body, making it easier for tumor cells to escape from the immune monitoring and attack of the body (22). During the treatment, the elderly patients may not be able to tolerate the standard dose of radiotherapy and chemotherapy, thus affecting the therapeutic effect and leading to an increased risk of adverse clinical outcomes. Therefore, when formulating treatment plan, we need to fully consider the age and physical condition of elderly patients and carry out individualized treatment.

Although the Nomogram prediction model constructed in this study showed good calibration degree and prediction performance in the training set and validation set, it still had certain limitations. First, all the samples in this study were from the same hospital, and there might be a certain selection bias. Patients in different regions and hospitals have differences in race, living habits, and medical level, which may affect the universality of the model. Second, while our sample size met minimum requirements for predictive modeling, the single-center derivation may affect generalizability. Formal power calculations were not performed for outcome prediction; future multicenter studies with pre-specified sample sizes based on expected event rates will be valuable to confirm our findings. Third, no external validation has been performed, making the applicability of the model in other clinical settings unclear. However, the sample size of this study was relatively limited, and external validation required a large amount of time, manpower and material resources, and involved a complex process of multi-center collaboration, which faced great difficulties in the development of this study, which was the main reason for the lack of external validation. Large-scale, multi-center external validation is a critical step for enhancing and demonstrating the generalizability and clinical utility of machine learning models. External validation using multi-center cohorts with diverse demographic and clinical characteristics is essential to confirm our findings. In addition, the lack of systematic data on perioperative variables (e.g., surgical approach (laparoscopic vs. open), extent of lymphadenectomy, intraoperative blood loss, and anesthesia techniques) may introduce unmeasured confounding. Future multicenter studies with standardized perioperative data collection are needed to validate and refine our model by incorporating these clinically relevant parameters.

In this study, we successfully constructed a Nomogram prediction model for the clinical outcome of postoperative gastric cancer in the elderly based on multiple factors such as NLR, BMI, tumor size, lymph node metastasis, CEA, and age. The model has good calibration degree and prediction performance in the training set and the validation set, and has certain clinical application value. It can provide a reference for clinicians to predict the clinical outcome of elderly patients with gastric cancer after surgery, and help to develop personalized treatment options. However, since no external validation is conducted in this study, the universality of the model needs to be further verified. In future studies, we should expand the sample size and carry out multi-center research to further improve and verify the model, so as to better serve the clinical diagnosis and treatment of elderly patients with gastric cancer and improve the survival rate and quality of life of patients.