Prognostic factors for overall survival in patients with early gastric cancer: a retrospective cohort study based on SEER database

Introduction

Gastric cancer (GC) remains a major global health concern, predictions for 2024 indicate that there will be over 960,000 new cases of GC and nearly 660,000 deaths worldwide (1,2), making it the fifth most common cancer and the third leading cause of cancer-related deaths in global (3). Early detection is pivotal—5-year survival exceeds 90% for early GC (EGC) but drops below 30% once the disease is advanced (4). The depth of tumor invasion and lymph node metastasis (LNM) are generally considered the most important factors for predicting the prognosis of GC (5,6). Age is also a predictive factor for various types of cancers (7), and surveys indicate that the incidence of stomach cancer increases with age, peaking at 60–70 years old (8). Therefore, analyzing the relationship between age and survival rates in GC can help elucidate the prognostic significance of age and potentially improve therapeutic outcomes. Sui et al. (9) pointed out that gender is a predictive factor for LNM; however, few studies have assessed the impact of both age and gender on the prognosis of EGC. Meanwhile, the influence of tumor size on the prognosis of EGC has rarely been reported. This study aimed to investigate the predictive factors of EGC using data from the Surveillance, Epidemiology, and End Results (SEER) database (a substantial tumor registry managed by the National Cancer Institute in the United States) (10), while also evaluating the influence of tumor size on the prognosis of EGC. Focusing on EGC can significantly improve patients’ survival rates and quality of life. For EGC, in-depth research on risk factors affecting prognosis can optimize treatment regimens, provide more precise treatment guidance for clinical practice, and promote the advancement of treatment technologies for EGC. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-405/rc).

Methods

Data source and patient selection

This is a population-based retrospective cohort study using data from the SEER program of the National Cancer Institute. In our study, we were authorized to use the SEER*Stat software (version v8.4.0.1; http://seer.cancer.gov/seerstat/) to download the information of 1,340 EGC patients who underwent surgical treatment and were registered in the SEER database from 2010 to 2015. EGC was defined as GC confined to the mucosa or submucosa regardless of LNM status. The inclusion criteria used in this study were as follows: (I) patients diagnosed with differentiated GC between 2010 and 2015, aged 18 years or older; (II) surgery was performed for histologically confirmed GC; (III) the staging of was Tis or T1, that is EGC; and (IV) had valid follow-up data. Exclusion criteria included: (I) missing data of Clinical pathological characteristics; (II) uncertain pathological stage or tumor location; (III) presence of distant metastasis prior to surgery; and (IV) lack of active follow-up or survival status. Finally, 1,340 patients with EGC were enrolled in the primary cohort (Figure 1). This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Figure 1 Flow diagram of sample selection and model construction. SEER, Surveillance, Epidemiology, and End Results.

The follow-up data in this study were derived from the survival status records in the SEER database, and the follow-up information was obtained through regular matching with the National Death Index (NDI) and the death registration systems of various states in the United States. The primary prognostic outcome was overall survival (OS), defined as the interval from the date of surgery to the date of death from any cause or date of last follow-up visit.

Variable definitions

Included variables were age (<60 and ≥60 years), gender (male and female), race (White, Black, and other), primary site (fundus of stomach, body of stomach, gastric antrum and pylorus, lesser curvature, greater curvature, overlapping, and stomach, not otherwise specified), histology (adenocarcinoma, mucinous adenocarcinoma, and signet ring cell carcinoma), degree of differentiation (well differentiation, moderate differentiation, poorly differentiation, and undifferentiation), depth of invasion (T1a and T1b), LNM (yes and no), and the number of dissected lymph nodes (<16 and ≥16). In addition, X-tile software was used to determine the optimal cutoff points for the tumor size variable in this study.

Statistical analysis

All statistical analyses and survival probability charts were created using SPSS 26.0 (IBM Corporation, Armonk, NY, USA) and R software (version 4.2.1). The X-tile version 3.6.1 software was employed to analyze and determine the optimal cut-off point for tumor size. The chi-square test was used for categorical data analysis. The Kaplan-Meier method was applied to estimate the OS rates among groups with different tumor sizes. The Cox proportional hazards regression model was utilized to identify prognostic factors for OS, model construction adopted the “backward stepwise method” (likelihood ratio test), with the inclusion criterion set as P<0.05 and the exclusion criterion set as P>0.10. The variables examined included age, gender, race, primary site, differentiation grade, numbers of retrieved lymph nodes, histology, LNM, depth of tumor invasion. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. A two-sided P value <0.05 was considered statistically significant.

Results

Clinicopathological characteristics

A total of 1,340 EGC patients who had undergone surgical treatment were retrieved from the SEER database. According to the patients’ tumor diameters and survival rates, the X-tile software was used to determine that 20 mm was the optimal cut-off point for tumor size, as shown in Figure 2. Based on this cut-off point, the patients were divided into two groups: the small-size tumors (SST) group, which consisted of 741 patients (55.3%) with a tumor diameter ≤20 mm, and the large-size tumors (LST) group, which included 599 patients (44.7%) with a tumor diameter >20 mm. The 5-year survival rates of the two groups were 64.9% and 52.8%, respectively, with statistical significance (P<0.001). as shown in Figure 3. The clinical and pathological data of both groups are presented in Table 1.

Figure 2 Map of the frequency distribution of tumor size in patients. No., number.

Figure 3 Comparison of survival curves between the large diameter group and the small diameter group. Large-size tumors: tumor diameter of >20 mm; small-size tumors: tumor diameter of ≤20 mm.

Analysis of prognostic factors for EGC

A univariate analysis was meticulously conducted on the clinical information of the patients. The findings revealed that gender (P<0.001), age (P<0.001), race (P<0.001), the quantity of dissected lymph nodes (P<0.001), LNM status (P<0.001), histological type (P=0.001), invasion depth (P<0.001), and tumor size (P<0.001) were all factors significantly associated with the prognosis of all patients, as delineated in Table 2. Subsequently, the Cox proportional hazards model analysis demonstrated that gender (P<0.001), age (P<0.001), race (P<0.001), the number of dissected lymph nodes (P=0.003), LNM (P<0.001), invasion depth (P=0.002), and tumor size (P<0.001) emerged as independent risk factors with a notable impact on the prognosis of EGC patients, as illustrated in Table 3.

Analysis of prognostic influencing factors in different age groups

The patients were categorized into two distinct groups based on age, namely the group with patients <60 years old and the group with patients ≥60 years old. Subsequently, the clinical information of these patients was analyzed using the Cox proportional hazards model, as presented in Table 4. It is noteworthy that for these different age-based groups, the prognostic influencing factors of EGC exhibit notable variations.

In terms of gender, the survival rate of female patients is better than that of male patients. Regarding race, the prognosis of patients from other ethnic groups is better than that of White and Black patients. As for the number of dissected lymph nodes, it is not an independent risk factor for prognosis among patients younger than 60 years old. However, among patients aged 60 years or older, a greater number of dissected lymph nodes is associated with a better prognosis, and it constitutes an independent risk factor.

In the context of LNM, patients without LNM exhibit a better prognosis compared to those with LNM. With regard to tumor size, among patients younger than 60 years old, tumor size does not serve as an independent prognostic risk factor. Conversely, among patients aged 60 years or older, a larger tumor size is associated with a poorer prognosis, and it is identified as an independent prognostic risk factor.

Regarding the depth of invasion, among patients younger than 60 years old, the depth of invasion does not significantly affect prognosis. Conversely, among patients aged 60 years or older, a greater depth of invasion is associated with a poorer prognosis.

Analysis of prognostic influencing factors in different gender groups

Multivariate Cox analysis based on patients’ different genders revealed that (Table 5), for different genders, the prognostic influencing factors of EGC vary. For both males and females, age, race, LNM, and tumor size are all independent risk factors affecting prognosis. Regarding the number of dissected lymph nodes, in male patients, the number of dissected lymph nodes does not significantly influence prognosis. However, in female patients, a greater number of dissected lymph nodes is associated with a better prognosis.

In terms of the depth of invasion, among male patients, a greater depth of tumor invasion is associated with a poorer prognosis. Conversely, among female patients, the depth of invasion does not have a significant impact on prognosis.

Stratified analysis of independent risk factors for patient prognosis

After conducting a stratified analysis of the number of dissected lymph nodes, LNM status, and depth of invasion, it was found that, among EGC patients, regardless of whether the number of dissected lymph nodes was <16 or ≥16, and whether the depth of invasion was the inner mucosa or the submucosa, the survival rate of the small-diameter group was significantly higher than that of the large-diameter group. For EGC patients without LNM, the survival rate of the large-diameter group was significantly lower than that of the small-diameter group (P<0.001), as presented in Table 6.

Discussion

EGC often presents with no obvious symptoms or only mild symptoms, and many patients are already in the advanced stage at the time of diagnosis (11). Understanding the risk factors for the prognosis of EGC is crucial for the early identification of high-risk individuals, allowing for the implementation of corresponding treatment measures. This is of great significance for improving patients’ prognosis. Tumor size serves as a reliable prognostic factor for GC patients, and the measurement of tumor size is beneficial for the staging and management of GC (12). Currently, there is no universally recognized method for selecting the cut-off point of tumor size. Commonly used approaches include using the receiver operating characteristic (ROC) curve to determine the optimal cut-off point of tumor size. In contrast, X-tile software represents a novel method. It obtains the optimal cut-off value through transient cut-off value analysis based on survival information and then identifies the cut-off value with the minimum P value from the log-rank test statistic of the survival rate of the classification biomarker (13). Compared with the ROC curve method, which relies solely on outcomes, using X-tile software appears to be more appropriate. Therefore, calculations using X-tile software revealed that when the cut-off point of tumor diameter was set at 20 mm, there was a significant difference in the survival rate between the two groups of patients. Thus, 20 mm can be regarded as the optimal cut-off point for tumor size in this group of patients.

From current GC research, it is evident that multiple prognostic factors can influence the survival rate of GC patients (14). Among these factors, the prognostic value of tumor size is often overlooked, and there is a lack of consensus on the optimal cut-off point for tumor size. A substantial number of studies have concentrated on the prognostic value of tumor size in GC (15-17), and it has been confirmed that tumor size is a non - negligible prognostic factor for GC, which can enhance the accuracy of survival prediction (15,18,19). The present study demonstrates that when the cut-off point of tumor diameter is set at 20 mm, there is a significant difference in the survival rate between the small-diameter group and the large-diameter group, and this difference is statistically significant. Further multivariate Cox regression analysis revealed that, in line with previous research findings, tumor size is also an independent prognostic factor for the patients in this group (15,20). Additionally, among EGC patients, the predictive ability of tumor size surpasses that of many other widely used prognostic factors. Factors such as LNM and depth of invasion also impact the relationship between tumor diameter and prognosis (21,22). Through stratified analysis of age, LNM and depth of invasion in this study, it was found that for EGC patients without LNM or those aged ≥60 years, tumor size can significantly affect their prognosis. Conversely, for EGC patients aged <60 years, tumor size does not influence their prognosis. This may be attributed to the fact that patients aged <60 years are more prone to LNM. The research results of Yin et al. (23) indicate that the probability of LNM in patients aged <60 years is more than five times that of patients aged ≥60 years. For EGC patients with positive lymph nodes, the correlation between tumor size and local metastasis is relatively weak; thus, tumor size does not affect their prognosis. For EGC patients without LNM, a smaller tumor size is associated with a better prognosis. Regarding the depth of invasion, regardless of whether the tumor is located in the inner mucosa or the submucosa, a smaller tumor size results in a smaller direct contact area with surrounding normal tissues, leading to a better prognosis. Both univariate and multivariate Cox regression analyses in this study showed that tumor size is one of the prognostic factors. This indicates that tumor size is a reliable indicator for assessing the biological behavior of EGC and improves the accuracy of patient prognosis prediction.

Age is also one of the crucial prognostic factors for many cancers (24). The prognosis of GC differs with age, and the survival rate of young GC patients is higher than that of elderly patients. Many studies have compared the long-term prognosis of GC between elderly patients and young or middle-aged patients. Most of these studies suggest that there is a difference in long-term survival between elderly patients and young or middle-aged GC patients (25). Our study has also confirmed the previous research findings. The prognosis of patients aged ≥60 years is worse than that of patients aged <60 years. This might be associated with the better physical condition of young patients. Young patients have an advantage over elderly patients in terms of immunity, surgical tolerance, and tolerance to chemotherapy drugs. Moreover, as people age, their bodily functions deteriorate. Therefore, more attention should be paid to elderly patients with EGC. There are relatively few studies on the impact of different genders on the prognosis of EGC. Han et al. (26) analyzed 122,793 GC patients, there were significantly more male patients (60.5%) than female patients (39.5%), and sex was an independent risk factor for prognosis. Li et al. (27) investigated 96,501 GC patients, among whom 61,639 were male and 34,862 were female. The report indicated that male patients had larger tumors, higher stages, and more advanced grades of GC. Additionally, the survival rate of female patients was significantly higher. Another study showed that the survival rate of elderly men was worse than that of elderly women, as they usually had more comorbidities than women of the same age (28). This result is consistent with that of our study. The survival rate of female patients with EGC is higher than that of male patients.

This study also has certain limitations. First, being a retrospective study, it sources patient data from the SEER database, which may lead to certain biases in data selection. Second, the enrolled patients cover a wide time span. With the continuous update of the comprehensive diagnosis and treatment model for GC, there are differences in the extent of radical surgery, the scope of lymph node dissection, and the content of pathological reports. Third, preoperative imaging data, tumor markers, and other relevant data are missing, preventing their inclusion in this study for analysis. Therefore, the main focus of our research is on analyzing the impact of age, gender, and tumor size on survival. To verify our findings, we plan to construct a large-sample, multi-center prospective study in the future to enhance our research.

Conclusions

An analysis of the SEER database has identified that age, gender, race, the number of dissected lymph nodes, LNM status, depth of invasion, and tumor size are all independent factors influencing the prognosis of EGC patients. Moreover, further stratified analysis reveals that these prognostic risk factors vary significantly across different patient subgroups. Therefore, in clinical practice, it is crucial to comprehensively assess these diverse clinical characteristics. Such an approach can offer more valuable reference information for evaluating patients’ conditions accurately, formulating personalized treatment plans, and predicting their prognosis effectively.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-405/rc

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-405/prf

Funding: This work was supported by the Tianjin Medical Key Discipline (Specialty) Construction Project (No. TJYXZDXK-058B) and the Tianjin Health Research Project (No. TJWJ2023XK014).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-405/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Thrift AP, Wenker TN, El-Serag HB. Global burden of gastric cancer: epidemiological trends, risk factors, screening and prevention. Nat Rev Clin Oncol 2023;20:338-49. [Crossref] [PubMed]
2. Sundar R, Nakayama I, Markar SR, et al. Gastric cancer. Lancet 2025;405:2087-102. [Crossref] [PubMed]
3. Zhang X, Yu X, Shen Q, et al. The role of TMSB15A in gastric cancer progression and its prognostic significance. J Gastrointest Oncol 2025;16:27-40. [Crossref] [PubMed]
4. Fu XY, Mao XL, Chen YH, et al. The Feasibility of Applying Artificial Intelligence to Gastrointestinal Endoscopy to Improve the Detection Rate of Early Gastric Cancer Screening. Front Med (Lausanne) 2022;9:886853. [Crossref] [PubMed]
5. Yanzhang W, Guanghua L, Zhihao Z, et al. The risk of lymph node metastasis in gastric cancer conforming to indications of endoscopic resection and pylorus-preserving gastrectomy: a single-center retrospective study. BMC Cancer 2021;21:1280. [Crossref] [PubMed]
6. Bozkurt HA, Ercan LD, Guden I, et al. Morphological predictors of lymph node metastasis in early gastric cancer. BMC Surg 2025;25:287. [Crossref] [PubMed]
7. Choi Y, Kim N, Kim JH, et al. Survival Rates of Patients with Gastric Cancer According to Age and Sex: A Large-Scale Study Using Data from 14,739 Patients. Cancer Res Treat 2025; Epub ahead of print. [Crossref]
8. Song P, Wu L, Jiang B, et al. Age-specific effects on the prognosis after surgery for gastric cancer: A SEER population-based analysis. Oncotarget 2016;7:48614-24. [Crossref] [PubMed]
9. Sui W, Chen Z, Li C, et al. Nomograms for Predicting the Lymph Node Metastasis in Early Gastric Cancer by Gender: A Retrospective Multicentric Study. Front Oncol 2021;11:616951. [Crossref] [PubMed]
10. Hankey BF, Ries LA, Edwards BK. The surveillance, epidemiology, and end results program: a national resource. Cancer Epidemiol Biomarkers Prev 1999;8:1117-21.
11. Sun Y, Yan F, Yu X, et al. Retrospective study on the prognostic prediction of inflammatory markers and the C-reactive protein/albumin ratio in first-line immunotherapy for advanced HER2 negative gastric cancer patients. Transl Cancer Res 2025;14:2043-53. [Crossref] [PubMed]
12. Xiao X, Gao B, Pang S, et al. Tumor size as a significant prognostic factor in T1 gastric cancer: a Surveillance, Epidemiology, and End Results (SEER) database analysis. BMC Gastroenterol 2023;23:121. [Crossref] [PubMed]
13. Camp RL, Dolled-Filhart M, Rimm DL. X-tile: a new bio-informatics tool for biomarker assessment and outcome-based cut-point optimization. Clin Cancer Res 2004;10:7252-9. [Crossref] [PubMed]
14. Wu YC, Cai YL, Rong L, et al. Characteristics of lymph node metastasis and evaluating the efficacy of endoscopic submucosal dissection in early gastric cancer. Beijing Da Xue Xue Bao Yi Xue Ban 2020;52:1093-7. [Crossref] [PubMed]
15. Zhou L, Li W, Cai S, et al. Large tumor size is a poor prognostic factor of gastric cancer with signet ring cell: Results from the surveillance, epidemiology, and end results database. Medicine (Baltimore) 2019;98:e17367. [Crossref] [PubMed]
16. Deng J, Zhang R, Pan Y, et al. Tumor size as a recommendable variable for accuracy of the prognostic prediction of gastric cancer: a retrospective analysis of 1,521 patients. Ann Surg Oncol 2015;22:565-72. [Crossref] [PubMed]
17. Xiao J, Shen K, Fan H, et al. Prognostic Value of Tumor Size in Gastric Cancer: A Retrospective Cohort Study Based on SEER Database. Int J Surg Pathol 2023;31:1273-82. [Crossref] [PubMed]
18. Peng C, Li T. A prognostic nomogram utilized lymph node ratio for signet ring cell gastric cancer patients post-surgery. J Gastrointest Oncol 2025;16:950-64. [Crossref] [PubMed]
19. Chen Y, Jia Y, Peng Z, et al. The prognostic role of tumor size in stage T1 gastric cancer. World J Surg Oncol 2022;20:135. [Crossref] [PubMed]
20. Del Rio P, Viani L, Bertocchi E, et al. The prognostic role of tumor size in patients with gastric cancer. Ann Ital Chir 2017;88:478-84.
21. Pang S, Wang W, Zhou J, et al. Inverse association between age and risk of lymph node metastasis in patients with early gastric cancer: a surveillance, epidemiology, and end results analysis. J Cancer 2024;15:2829-36. [Crossref] [PubMed]
22. Oh SE, Ahn S, Kim KM, et al. Identification of maximal tumor size associated with negligible lymph node metastasis for endoscopic submucosal dissection of undifferentiated-type early gastric cancer. Gastric Cancer 2024;27:850-7. [Crossref] [PubMed]
23. Yin J, Cai J, Zhang J, et al. Study on related risk factors and prognosis for lymph node metastasis in patients with early gastric cancer. Chinese Journal of Practical Surgery 2016;36:111-4, 118.
24. Sarriugarte A, Escaño EA, Valero AP, et al. Implications of the age in surgical treatment results and prognosis of gastric cancer. Eur J Surg Oncol 2022;48:e183-4.
25. Li XB, Mei LY, Che XM. Prognostic factors for gastric cancer. China Journal of General Surgery 2013;22:494-7.
26. Han D, Zhang Z, Deng J, et al. Survival analysis of gastric malignancy patients: identifying key prognostic factors. Transl Cancer Res 2025;14:1928-41. [Crossref] [PubMed]
27. Li H, Wei Z, Wang C, et al. Gender Differences in Gastric Cancer Survival: 99,922 Cases Based on the SEER Database. J Gastrointest Surg 2020;24:1747-57. [Crossref] [PubMed]
28. Lee JY, Kim HI, Kim YN, et al. Clinical Significance of the Prognostic Nutritional Index for Predicting Short- and Long-Term Surgical Outcomes After Gastrectomy: A Retrospective Analysis of 7781 Gastric Cancer Patients. Medicine (Baltimore) 2016;95:e3539. [Crossref] [PubMed]