A prognostic nomogram based on log odds of positive lymph nodes to predict survival in gastric cancer patients with neoadjuvant chemotherapy: a large population-based cohort study and external validation

Introduction

Gastric cancer (GC) is one of the most prevalent malignancies globally, ranking fifth in terms of incidence and third in cancer-related mortality (1). Although the overall incidence of GC is declining, it still accounts for over one million new cases each year, with a particularly high prevalence in East Asia (2). Early-stage GC often presents with few distinct clinical symptoms, resulting in most patients being diagnosed at an advanced stage. Advanced GC is associated with a high recurrence rate and poor prognosis, with a median overall survival (OS) of less than 12 months (3). Surgery, chemotherapy, radiotherapy, immunotherapy, and targeted therapy have all been recognized as beneficial for patients with advanced GC (4). However, radical resection (also known as R0 resection) remains the only potentially curative treatment for advanced GC. Therefore, increasing the rate of radical resection and minimizing postoperative recurrence are key priorities in GC therapy.

In recent years, neoadjuvant chemotherapy has been increasingly used in the treatment of GC and is now recommended by several clinical guidelines (5). Theoretically, neoadjuvant chemotherapy can reduce recurrence rates and prolong survival (6). Many studies have focused on identifying patients at risk for poor prognosis, with the aim of intervening and potentially improving outcomes.

According to the 8th edition of the Union for International Cancer Control (UICC) and the American Joint Committee on Cancer (AJCC) tumor-node-metastasis (TNM) staging system, the status of regional lymph nodes is a key prognostic factor in patients with GC (7). Guidelines recommend that at least 16 regional lymph nodes be retrieved during surgery, with a preference for more than 30 (8-10). The lymph node ratio (LNR), defined as the percentage of positive lymph nodes (PLNs) among the examined lymph nodes (ELNs), has been reported as a potentially valuable prognostic predictor in patients with gastric, colon, lung, and breast cancers (11-15). Log odds of positive lymph nodes (LODDS), defined as the logarithm of the ratio of PLNs to negative lymph nodes (NLNs), have been reported by several studies to have equal or even superior predictive ability compared to LNR in patients with partial GC (16,17). However, the predictive abilities of LODDS and LNR in GC patients who receive systemic therapy before surgery remain a topic of debate.

In this study, we aimed to explore the potential value of various clinical features (including but not limited to LNR and LODDS) in predicting the outcomes of GC patients who receive systemic therapy before surgery, and to establish and validate a nomogram for forecasting these outcomes. We present this article in accordance with the TRIPOD reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1013/rc).

Methods

Data sources

We obtained data from the Surveillance, Epidemiology, and End Results (SEER) research database (November 2020 Submission) (18) using SEER*Stat software version 8.4.0 (http://seer.cancer.gov/seerstat/). Information on incidence-SEER Research Plus Data, 18 registries, Nov 2019 Sub (2000–2019) was tested.

We also collected data from GC patients as an external validation group, who were treated at The First Affiliated Hospital of Wenzhou Medical University between January 2015 and September 2022, with the last follow-up occurring on September 22, 2025. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The First Affiliated Hospital of Wenzhou Medical University (Approval No. [KY2023-R229]) and individual consent for this retrospective analysis was waived. The authors also signed the Data Use Agreement for the SEER database.

Participants of inclusion

Inclusion criteria: (I) patients aged 18 to 80 years; (II) diagnosis of GC confirmed by positive histology; (III) complete clinical features (including but not limited to age, sex, ethnicity, radiation treatment, and number of lymph nodes) and follow-up data available; (IV) patients who underwent systemic chemotherapy before surgery; (V) patients with stage I–III disease. Exclusion criteria: (I) patients who only had autopsies or death certificates available; (II) patients whose survival time was less than 1 month; (III) patients with unclear sequences of systemic therapy, radiation therapy, and surgery; (IV) patients with two or more primary cancers; (V) patients who received radiation treatment before surgery; (VI) patients who did not undergo surgery.

Variates and definitions

Age was categorized into three groups (<45, 45–60, and >60 years) based on a previous study (19). Sex was categorized into male and female. The primary site recode information, based on the criteria in the 2nd edition of the International Classification of Diseases for Oncology (ICD-O-2), was divided into seven categories: cardia, fundus of the stomach, body of the stomach, gastric antrum, pylorus, lesser curvature of the stomach, greater curvature of the stomach, and overlapping lesions of the stomach. Based on the criteria in the ICD-O-2, cases with histologic codes from 8140 to 8389 were classified as adenocarcinoma, those with codes 8480 to 8481 were classified as mucinous adenocarcinoma/mucin-producing adenocarcinoma (AM/MPA), and those with code 8490 were classified as signet ring cell carcinoma (SRCC). Histologic differentiation grade was categorized into four groups: grade I, grade II, grade III, and grade IV. The number of NLNs was calculated using the formula: NLNs = ELNs − PLNs. LNR was the ratio of PLNs to ELNs, which has been shown to be a potential predictor for the outcome of several cancers (11-14). LODDS was defined as the log of the ratio between the number of PLNs and the number of NLNs, calculated using the formula: LODDS=log[(PLNs + 0.5)/(NLNs + 0.5)] (20). The cut-off values for LNR, ELNs, and NLNs were determined using the Kaplan-Meier method. This method involves performing a log-rank test at different cut-off points to identify the value that best separates the survival curves into distinct groups with significant differences. LNR, ELNs, and NLNs were divided into two groups each according to their respective cut-off values. LODDS was categorized into three levels based on tertiles, as described by Lee et al. in their research: LODDS1 (<−1.430), LODDS2 (−1.430 to −0.639), and LODDS3 (>−0.639) (21). Survival months were calculated using the following formula: survival months = FLOOR [(endpoint − date of diagnosis) / days in a month], based on its definition in the SEER database (https://seer.cancer.gov/survivaltime/). OS was defined as the time from diagnosis to death.

Risk factors exploration

All patients in the SEER database were randomly assigned to a development group and an internal validation group in a 7:3 ratio. The chi-square test or Fisher’s exact test was used to compare the characteristics of patients between the training and validation datasets. Independent factors associated with the OS of GC patients receiving preoperative systemic chemotherapy were initially screened using univariate and multivariate logistic regression analysis. The least absolute shrinkage and selection operator (LASSO) regression algorithm was used to find out the most relevant features from the statistically significant features in univariate logistic analysis. To find out the most suitable tuning parameters (λ) for LASSO-logistic regression, 10-fold cross-validation was done, and the most significant features were screened out. Odds ratios (ORs) and 95% confidence intervals (CIs) were used to quantify the effects of variables on OS. A generalized linear model was built based on the results of the logistic regression analysis. To visually display the generalized linear model, a forest plot was created. To assess the predictive accuracy of the model, receiver operating characteristic (ROC) curve analysis was performed, and the area under the curve (AUC) was calculated for both the development and validation groups. The AUC values ranged from 0.5 to 1.0, demonstrating the reliability of the generalized linear model. An AUC greater than 0.7 is generally considered good, while an AUC above 0.8 is considered excellent. Based on the results of univariate and multivariate logistic regression analysis, univariate and multivariate Cox regression analyses were subsequently performed. Hazard ratios (HRs) and 95% CIs were used to evaluate the effects of features in Cox regression analysis. Schoenfeld’s global test (22) proceeded to verify whether features conform to the proportional hazard (PH) assumption. Deviance residual diagrams were used to evaluate the distribution of data in each variate.

Nomogram construction and validation

Based on the results of the analyses above, a nomogram was constructed to predict the 1-, 3-, and 5-year OS of GC patients receiving preoperative systemic chemotherapy. The concordance index (C-index) of the nomogram was assessed separately in the training and validation datasets. To assess the predictive accuracy of the nomogram, 1-, 3-, and 5-year ROC analyses were performed, and AUCs were calculated for both the development and validation groups. Validation curves for 1-, 3-, and 5-year OS in both datasets were generated using a 1,000-resample bootstrapping method. Based on the training set data, 1-, 3-, and 5-year decision curve analyses (DCA) were constructed to evaluate the practical application value of the nomogram and AJCC TNM staging tools in clinical decision-making.

Statistical analysis

In this study, categorical variables were coded numerically and tested using the chi-square test or Fisher’s exact test to compare features between the development and validation groups. The relationships between clinical and pathological variables and OS were explored using univariate and multivariate logistic regression, as well as univariate and multivariate Cox regression analysis. Independent risk factors for GC patients receiving preoperative systemic chemotherapy were identified using these methods. The LASSO algorithm was applied to select the most relevant features from those statistically significant in the univariate logistic analysis. To assess the predictive performance of the model, C-index, ROC, and AUC were calculated for both the development and validation groups. Calibration curves were plotted to visually assess model accuracy. Cox PHs regression was performed to examine the impact of features on OS in both univariate and multivariate models. The PHs assumption was tested using Schoenfeld’s global test, and deviance residual plots were used to evaluate model fit. Nomogram construction was performed based on the results of regression analyses to predict 1-, 3-, and 5-year OS. The C-index, ROC analyses, and DCA were used to evaluate the nomogram’s predictive accuracy, and validation curves were generated through bootstrapping with 1000 resamples. All analyses and figures were performed using R software (version 4.1.2, https://www.r-project.org/). R packages “survival” and “survminer” were used for survival analysis and visualization of survival curves, “caret” for data partitioning, model training, and validation, “tableone”, “survival”, and “survminer” for creating summary tables of baseline characteristics, “tidyverse” for data manipulation and visualization, “nnet” for multinomial logistic regression, “compareGroups” for comparing groups in statistical analysis, “rmda” and “ggDCA” for DCA, “glmnet” for LASSO regression and regularized logistic regression, “rms” for regression modeling strategies, including nomogram construction, “plyr” for data manipulation, “forestplot” for creating forest plots from model results, “timeROC” and “pROC” for time-dependent ROC analysis, and “ezcox” for Cox regression with time-dependent covariates. Statistical significance was defined as two-sided P values <0.05.

Results

Characteristics of patients identified

The process of patient identification is displayed in Figure 1. A total of 726 cases were downloaded from the SEER database, and 124 cases were identified from the database of The First Affiliated Hospital of Wenzhou Medical University. Cutoff values for LNR, ELN, and NLN were calculated using the Kaplan-Meier method and are shown in Figure 2A-2C. Patients were divided into two LNR-related groups: low (≤0.15) and high (>0.15), two NLN-related groups: low (≤17) and high (>17), and two ELN-related groups: low (≤25) and high (>25), respectively. The Kaplan-Meier curves for LNR-related groups, NLN-related groups, and ELN-related groups are shown in Figure 2D-2F. The Kaplan-Meier curves indicated significant differences in survival between the LNR-related groups, NLN-related groups, and ELN-related groups. All SEER cases were randomly divided into the training dataset (510 cases) and the internal validation dataset (216 cases). Cases from our hospital were grouped as the external validation group. The details of cases between datasets are listed in Table 1. Nearly half of the patients were aged older than 60 years old (482 cases, 56.7%), stage III (468 cases, 55.1%), and T3 stage (394 cases, 46.4%). More than half of them were male (599 cases,70.5%), white people (494 cases, 58.1%), grade III (639 cases, 75.2%), adenocarcinoma (641 cases, 75.2%), low LNR (547 cases, 64.4%), low ELN (573 cases, 67.4%), high NLN (425 cases, 50.0%). In this study, 451 cases (53.1%) were alive at the end of follow-up, while 399 cases (46.9%) were dead. The median survival time of all patients was 33 months [interquartile range (IQR): 19.0–54.0 months].

Figure 1 The workflow of the patients chosen. AJCC, American Joint Committee on Cancer; M, metastasis; N, node; SEER, Surveillance, Epidemiology, and End Results; T, tumor; WMU, Wenzhou Medical University.

Figure 2 Identification of optimal cut-off values for lymph node parameters and corresponding overall survival analyses. The cut-off value of the number of lymph nodes examined (A), negative regional lymph nodes (B) and lymph node ratio (C). The survival analyses between different groups based on lymph nodes examined (D), negative regional lymph nodes (E) and lymph node ratio (F) were proceeded.

The exploration of independent risk factors for GC patients with systemic chemotherapy before surgery

Univariate and multivariate logistic regression analyses were performed on the training cohort to initially screen for independent risk factors for preoperative systemic chemotherapy or radiotherapy in patients with colorectal cancer. The LASSO regression algorithm was applied in this process. Based on LASSO-logistic analysis (Lambda.1SE =0.03833762, shown in Figure S1A,S1B), variables including primary site, T stage, N stage, and LODDS were selected (all P values <0.05). The results of univariate logistic regression and LASSO-logistic regression analyses are shown in Table 2. Compared with patients whose primary site was the cardia, patients whose primary site was the fundus of the stomach (P=0.01) and patients whose primary site was the greater curvature of the stomach (P=0.008) had better outcomes. Compared with patients with T1 stage, patients with T3 stage (P=0.002) and patients with T4 stage (P<0.001) had worse outcomes. Compared with patients with N1 stage, patients with N3 stage (P=0.04) had worse outcomes. Compared with patients with LODDS1, patients with LODDS2 (P=0.002) had worse outcomes. The same trends were observed in the external validation cohort, with LODDS2 (P=0.026) and LODDS3 (P=0.007) showing significant associations, as presented in Table S1. The corresponding LASSO regression process is illustrated in Figure S2A,S2B. Based on the variables identified, a generalized linear model was constructed and visualized in Figure 3A. ROC curves were used to estimate the reliability of the model, and AUC values were calculated for the construction, internal validation, and external validation groups. ROC curves and AUC values are shown in Figure 3B-3D. The AUC values for the construction, internal validation, and external validation groups were 0.830, 0.704, and 0.750, respectively, indicating that the model demonstrated a high degree of predictive capacity.

Figure 3 The LASSO-logistic regression analysis. (A) Forest plot of LASSO-logistic regression analysis. ROC curve analysis and the AUC in training cohort (B) internal validation cohort (C), external validation cohort (D). AUC, area under the curve; CI, confidence interval; LASSO, least absolute shrinkage and selection operator; LODDS, log odds of positive lymph nodes; N, node; OR, odds ratio; ROC, receiver operating characteristic; T, tumor.

Univariate and multivariate Cox regression analyses were performed for further exploration of independent risk factors. The results of the univariate and multivariate Cox regression analyses in train dataset are shown in Table 3, while the results of the univariate Cox regression analyses in external validation dataset are shown in Table S2. All features with significant differences in the univariate Cox regression analysis (P<0.05) were included in the multivariate Cox regression analysis. Schoenfeld’s global test was conducted and visualized in Figure S3A,S3B. According to the results of Schoenfeld’s global test, variables including primary site (P=0.21), T stage (P=0.059), N stage (P=0.35), and LODDS (P=0.43) conformed to the PHs assumption. Compared with patients whose primary site was the cardia, patients whose primary site was the body of the stomach (P=0.02), gastric antrum (P=0.005), and lesser curvature of the stomach (P=0.02) had better outcomes. Compared with patients with T1 stage, patients with T3 stage (P=0.01) and T4 stage (P=0.001) had worse outcomes. Compared with patients with N1 stage, patients with N3 stage (P<0.001) had worse outcomes. Compared with patients with LODDS1, patients with LODDS3 (P<0.001) had worse outcomes. The deviance residual plot indicated that the residuals for all variables included in the nomogram followed a symmetric pattern and had a consistent, uniform spread throughout the fit.

Nomogram construction and validation

Based on the features selected above, the nomogram was constructed and is shown in Figure 4. The C-index values for the nomogram in the training dataset, internal validation dataset, and external validation dataset were 0.755, 0.732, and 0.809, respectively. The 1-, 3-, and 5-year OS calibration curves for the development group, internal validation group, and external validation group are shown in Figure S4. The results of the 1-, 3-, and 5-year ROC analyses in the development group, internal validation group, and external validation group are shown in Figure 5A-5C. The 1-, 3-, and 5-year AUCs in the development dataset were 0.780, 0.790, and 0.803, respectively. The 1-, 3-, and 5-year AUCs in the internal validation dataset were 0.742, 0.730, and 0.752, respectively. The 1-, 3-, and 5-year AUCs in the external validation dataset were 0.747, 0.777, and 0.725, respectively. The calibration curves showed that the prediction model was consistent with the actual survival outcomes in both datasets. The plot of DCA based on the training set is shown in Figure 5D, revealing that, compared with the TNM staging, the nomogram constructed in this study demonstrated relatively high clinical application value.

Figure 4 Prognostic nomogram of gastric cancer patients treated with neoadjuvant chemotherapy based on 4 risk factors. LODDS, log odds of positive lymph nodes; N, node; T, tumor.

Figure 5 Reliability of nomogram. (A-C) 1-, 3-, 5-year receiver operating characteristic curve analysis and AUC in the training cohort (A), the internal validation cohort (B), and the external validation cohort (C). (D) 1-, 3-, 5-year decision curve analysis based on trainset were built. AJCC, American Joint Committee on Cancer; AUC, area under the curve; LODDS, log odds of positive lymph nodes; TNM, tumor-node-metastasis.

Discussion

In recent years, with the implementation of standardized surgery, adjuvant/neoadjuvant therapies (chemotherapy, radiotherapy), and the use of targeted or immunotherapy drugs in GC, the 5-year survival rate of GC patients has improved (23). Currently, the treatment of GC has evolved from simple surgery to a comprehensive approach based on multidisciplinary cooperation. Preoperative chemotherapy can reduce tumor size, facilitate downstaging, increase the likelihood of radical resection, and improve chemotherapy compliance (24). Many clinical trials, including MAGIC (25), FLOT (26), RESLOVE (27), have confirmed that neoadjuvant chemotherapy has a positive effect on improving the prognosis of patients with GC. Thus, neoadjuvant chemotherapy has become an important part of advanced GC patients’ treatment and is written into the guidelines. Currently, different guidelines provide varying descriptions of the appropriate population for neoadjuvant chemotherapy, but they all agree that neoadjuvant chemotherapy is beneficial for prognosis in selected patients with GC (4,28-30). Across the different guidelines, a common recommendation is that patients with lymphadenopathy are more likely to benefit from neoadjuvant chemotherapy.

Recent studies have demonstrated that GC exhibits distinct patterns of lymph node and distant metastasis compared with other malignancies (31). Therefore, various approaches, including bioinformatics methods, have been widely applied to investigate the mechanisms underlying GC development and progression, as well as its prognosis. For example, Gemmell et al. suggested that integrating traditional Chinese medicine with modern network pharmacology and omics sequencing technologies may provide potential benefits for the treatment of GC (32). It also indicates that traditional Chinese medicine has considerable potential and application prospects in the treatment of GC. A similar trend has also been observed in prognostic studies of GC, where an increasing number of indicators and analytical approaches have been introduced into this field. TNM staging is the most vital predictive factor for judging tumor progression, evaluating prognosis, and guiding follow-up treatment. The UICC/AJCC or the Japanese Gastric Cancer Association (JGCA) have formulated different staging standards for GC at different periods. In the following staging changes, the three major staging systems were gradually unified. The most important change during the period is the definition of N staging. Insufficient harvesting of lymph nodes can affect the accuracy of N staging and prognosis assessment (33,34). Smith et al. (35) found that regardless of lymph node metastasis status, using the number of detected lymph nodes as the grouping standard, the survival time of patients with detected lymph nodes ≥15 was better than that of patients with detected lymph nodes <15. Volpe et al. (36) analyzed 114 patients who underwent proximal radical gastrectomy (including D1, D1+, D2, and D2+), and found that the median survival improved from 25 to 42 months for patients who had 15 or more lymph nodes removed when treated with extended resection. The eighth edition of the TNM staging system recommends that the surgeon obtain at least 16 regional lymph nodes, preferably more than 30 (37). However, this guideline does not include the number of harvested lymph nodes in patients receiving neoadjuvant chemotherapy. The operator found that the lymph nodes become fibrotic or smaller after neoadjuvant chemotherapy, which affects the final detection of lymph nodes (38). Our study suggests that at least 25 lymph nodes should be obtained from these patients with GC, among which 17 NLNs are guaranteed, which will have a better prognosis. To the best of our knowledge, this finding has not been reported in previous studies and represents one of the novel aspects of the present study.

In the case of insufficient lymph node extraction, the method of LNR can effectively correct the stage migration caused by insufficient lymph node extraction to a certain extent. A prognostic analysis of 1,075 patients by Zhou et al. (39) pointed out that regardless of the number of lymph nodes examined, LNR staging was better than N staging in predicting the prognosis of patients and suggested that LNR staging replaced N staging to predict the status of lymph nodes. Many studies also support the addition of LNR to the TNM staging system, which can improve the limitations of N staging when the number of lymph nodes harvested is insufficient, and can accurately stratify the prognosis of GC patients (40,41). Therefore, LNR is more suitable for the judgment of the prognosis of GC when the number of lymph nodes dissected is insufficient in D1 dissection, but there is no formal regulation on the cut-off value of LNR, which limits its application in the world. The results of this study suggest that the optimal cut-off value of LNR is 0.15, which needs to be verified by subsequent larger clinical samples.

In recent years, some scholars have found the prognosis of GC patients varies with the number of resected lymph nodes, although the patients have the same LNR. In terms of prognosis assessment for patients with GC, Qiu et al. combined LODDS staging with T staging and distant metastasis, and put forward a new hypothesis, that is, the tumor-LODDS-metastasis (TLM) staging system can predict the prognosis of patients with accuracy (42). Sun et al. (43) analyzed 2,547 cases of GC patients undergoing D2 radical resection and found that LODDS is an independent prognostic factor, while LNR is not. They believed that LODDS is better than LNR in predicting the prognosis of GC patients, especially in the case of insufficient lymph node dissection, which is consistent with our findings.

In this study, we analyzed the prognosis of a total of 850 GC patients who received systemic chemotherapy before surgery and constructed a nomogram. LODDS can greatly supplement N staging with an insufficient lymph node dissection. Compared with LODDS1, patients in the LODDS3 subgroup (HR: 2.943, 95% CI: 1.843–4.703) had a worse outcome in OS. Consistent with our findings, multiple studies have demonstrated that the primary tumor site is an important independent prognostic factor in GC (44-46). What’s more, the TNM staging system is widely recognized as an effective tool for predicting prognosis in GC, as it incorporates tumor invasion depth, lymph node involvement, and distant metastasis. Based on these factors, we developed a prognostic model that demonstrated improved predictive performance compared with the TNM staging system for GC patients undergoing preoperative neoadjuvant chemotherapy. Our system may provide surgeons with additional guidance beyond conventional TNM staging. Several types of research have looked for characteristics that might be related to survival in GC patients undergoing preoperative neoadjuvant chemotherapy, however, most of these studies weren’t able to clarify how these factors might affect survival. To our knowledge, our nomogram is the first nomogram construction based on SEER and our central database for GC patients with neoadjuvant chemotherapy before surgery. In our research, LODDS and N staging were found to be independent risk variables both in logistic and COX regression analysis, while LNR didn’t show a significant association with the OS of GC patients we identified. The results of this study reveal that LODDS can be used as a supplement to N stage for GC patients who have received neoadjuvant chemotherapy before surgery, providing a certain reference for the prognosis of patients. LODDS was defined to enhance the accuracy of prognostic assessments, particularly for patients with N0 status or a total harvested LN count of fewer than 15 (47). Although the N stage and LODDS show similar trends in our model, their implications are distinct. LODDS incorporates both the number of ELNs and the number of NLNs, which helps account for variations in lymph node retrieval and reduces stage migration bias. Therefore, LODDS provides more refined stratification and additional prognostic information beyond standard N staging, particularly in patients with limited lymph node yield following neoadjuvant therapy.

There are still some limitations that we must consider for our nomogram. Firstly, the cut-off value of LODDS was decided according to the tertiles, just as Lee et al. did in their research (21). Therefor further exploration of the optimal cut-off value may be necessary for improving the prediction accuracy of the nomogram. Secondly, SEER is a large population-based database, the surgical details (extent of lymph node dissection, D1, D2, or D2+) were not available. Regarding the details of adjuvant treatment in this cohort of post-surgery patients, including whether they received postoperative therapy, we acknowledge that such information is not captured in the current SEER dataset. Therefore, caution is warranted when applying these findings to individual treatment decisions. These limitations further highlight the importance of future prospective, multi-center, and multi-ethnic studies incorporating comprehensive treatment and pathological information. Thirdly, the number of samples in the external validation group is limited, and the nomogram needs to be validated in a larger sample size. Our future studies are planned to be conducted in multi-center, multi-ethnic cohorts in order to obtain more reliable and generalizable results. Fourthly, although OS is widely used in survival analysis, it still has its shortcomings. For a better evaluation of the prediction ability of LODDS for GC patients, survival analysis based on patients’ progression-free survival (PFS) and cancer-specific survival (CSS) should be performed in the follow-up study.

Conclusions

In this study, we identified LODDS as a potential predictive factor for the prognosis of GC patients who undergo systemic therapy before surgery. All features included in the nomogram are readily available in clinical practice, making it a practical tool for clinicians. Our nomogram can serve as a valuable complement to the TNM staging system by providing additional predictive information. This combined approach allows for a more comprehensive and personalized treatment strategy, enhancing the accuracy of prognosis prediction for GC patients who receive neoadjuvant chemotherapy prior to surgery and influencing subsequent treatment decisions.

Acknowledgments

We acknowledge SEER database for providing their platforms and contributors for uploading their meaningful datasets.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1013/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1013/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1013/prf

Funding: This study was supported by the Industry-University-Research Innovation Fund for Chinese Universities (No. KJHX2313).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1013/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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of The First Affiliated Hospital of Wenzhou Medical University (No. KY2023-R229) and individual consent for this retrospective analysis was waived.

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. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Smyth EC, Nilsson M, Grabsch HI, et al. Gastric cancer. Lancet 2020;396:635-48. [Crossref] [PubMed]
3. Wang SY, Wang YX, Guan LS, et al. Construction of a prognostic model for gastric cancer based on immune infiltration and microenvironment, and exploration of MEF2C gene function. BMC Med Genomics 2025;18:13. [Crossref] [PubMed]
4. Wang FH, Zhang XT, Tang L, et al. The Chinese Society of Clinical Oncology (CSCO): Clinical guidelines for the diagnosis and treatment of gastric cancer, 2023. Cancer Commun (Lond) 2024;44:127-72. [Crossref] [PubMed]
5. Su PF, Yu JC. Progress in neoadjuvant therapy for gastric cancer. Oncol Lett 2022;23:172. [Crossref] [PubMed]
6. Joshi SS, Badgwell BD. Current treatment and recent progress in gastric cancer. CA Cancer J Clin 2021;71:264-79. [Crossref] [PubMed]
7. Kinami S, Saito H, Takamura H. Significance of Lymph Node Metastasis in the Treatment of Gastric Cancer and Current Challenges in Determining the Extent of Metastasis. Front Oncol 2021;11:806162. [Crossref] [PubMed]
8. Ajani JA, D'Amico TA, Bentrem DJ, et al. Gastric Cancer, Version 2.2025, NCCN Clinical Practice Guidelines In Oncology. J Natl Compr Canc Netw 2025;23:169-91. [Crossref] [PubMed]
9. Li W, Shao X, Ma F, et al. Total Lymph Node Yield Is Associated with Prolonged Survival after Neoadjuvant Chemotherapy in ypN0 Gastric Cancer Patients. Oncol Res Treat 2023;46:287-95. [Crossref] [PubMed]
10. Jiang Q, Zeng X, Zhang C, et al. Lymph node ratio is a prospective prognostic indicator for locally advanced gastric cancer patients after neoadjuvant chemotherapy. World J Surg Oncol 2022;20:261. [Crossref] [PubMed]
11. Shetiwy M, Elalfy AF, Eldamshety O, et al. The Prognostic Significance of Lymph Node Status and Lymph Node Ratio (LNR) on Survival of Right Colon Cancer Patients: a Tertiary Center Experience. J Gastrointest Cancer 2021;52:1010-5. [Crossref] [PubMed]
12. Sun Y, Li Y, Wu J, et al. Nomograms for prediction of overall and cancer-specific survival in young breast cancer. Breast Cancer Res Treat 2020;184:597-613. [Crossref] [PubMed]
13. Wei YL, Cui YH, Dong YH, et al. Zhonghua Yi Xue Za Zhi 2021;101:259-64. [Clinical study on the effects of neoadjuvant chemotherapy on lymph node ratio, prognosis, and survival analysis of stage IIIA-N2 non-small cell lung cancer]. [PubMed]
14. Ichhpuniani S, McKechnie T, Lee J, et al. Lymph Node Ratio as a Predictor of Survival for Colon Cancer: A Systematic Review and Meta-Analysis. Am Surg 2024;90:840-50. [Crossref] [PubMed]
15. Liu S, Da X, Shu L, et al. Prognostic value of lymph node ratio after neoadjuvant chemotherapy for gastric cancer: a systematic review and meta-analysis. Front Oncol 2026;16:1755605. [Crossref] [PubMed]
16. Pan S, Wang P, Xing Y, et al. Retrieved lymph nodes from different anatomic groups in gastric cancer: a proposed optimal number, comparison with other nodal classification strategies and its impact on prognosis. Cancer Commun (Lond) 2019;39:49. [Crossref] [PubMed]
17. Gu P, Deng J, Sun Z, et al. Superiority of log odds of positive lymph nodes (LODDS) for prognostic prediction after gastric cancer surgery: a multi-institutional analysis of 7620 patients in China. Surg Today 2021;51:101-10. [Crossref] [PubMed]
18. Paijens ST, Vledder A, de Bruyn M, et al. Tumor-infiltrating lymphocytes in the immunotherapy era. Cell Mol Immunol 2021;18:842-59. [Crossref] [PubMed]
19. Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
20. Baqar AR, Wilkins S, Wang W, et al. Log odds of positive lymph nodes is prognostically equivalent to lymph node ratio in non-metastatic colon cancer. BMC Cancer 2020;20:762. [Crossref] [PubMed]
21. Lee CW, Wilkinson KH, Sheka AC, et al. The Log Odds of Positive Lymph Nodes Stratifies and Predicts Survival of High-Risk Individuals Among Stage III Rectal Cancer Patients. Oncologist 2016;21:425-32. [Crossref] [PubMed]
22. Abeysekera WWM, Sooriyarachchi MR. Use of Schoenfeld’s global test to test the proportional hazards assumption in the Cox proportional hazards model: an application to a clinical study. J Natn Sci Foundation Sri Lanka 2009;37:41-5. [Crossref]
23. Allemani C, Matsuda T, Di Carlo V, et al. Global surveillance of trends in cancer survival 2000-14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. Lancet 2018;391:1023-75. [Crossref] [PubMed]
24. Fong C, Johnston E, Starling N. Neoadjuvant and Adjuvant Therapy Approaches to Gastric Cancer. Curr Treat Options Oncol 2022;23:1247-68. [Crossref] [PubMed]
25. Cunningham D, Allum WH, Stenning SP, et al. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 2006;355:11-20. [Crossref] [PubMed]
26. Al-Batran SE, Hofheinz RD, Pauligk C, et al. Histopathological regression after neoadjuvant docetaxel, oxaliplatin, fluorouracil, and leucovorin versus epirubicin, cisplatin, and fluorouracil or capecitabine in patients with resectable gastric or gastro-oesophageal junction adenocarcinoma (FLOT4-AIO): results from the phase 2 part of a multicentre, open-label, randomised phase 2/3 trial. Lancet Oncol 2016;17:1697-1708. [Crossref] [PubMed]
27. Zhang X, Liang H, Li Z, et al. Perioperative or postoperative adjuvant oxaliplatin with S-1 versus adjuvant oxaliplatin with capecitabine in patients with locally advanced gastric or gastro-oesophageal junction adenocarcinoma undergoing D2 gastrectomy (RESOLVE): an open-label, superiority and non-inferiority, phase 3 randomised controlled trial. Lancet Oncol 2021;22:1081-92. Erratum in: Lancet Oncol 2021;22:e347 Erratum in: Lancet Oncol 2025;26:e77. [Crossref] [PubMed]
28. Ajani JA, D'Amico TA, Bentrem DJ, et al. Gastric Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:167-92. [Crossref] [PubMed]
29. Lordick F, Carneiro F, Cascinu S, et al. Gastric cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol 2022;33:1005-20. [Crossref] [PubMed]
30. Japanese Gastric Cancer Treatment Guidelines 2021 (6th edition). Gastric Cancer 2023;26:1-25.
31. Lamprecht CB, Kashuv T, Lucke-Wold B. Understanding metastatic patterns in gastric cancer: Insights from lymph node distribution and pathology. World J Gastrointest Oncol 2025;17:103709. [Crossref] [PubMed]
32. Gemmell JS, Lucke-Wold B. Traditional Chinese medicine and modern technology: Network pharmacology and omics sequencing in gastric cancer. World J Gastrointest Oncol 2025;17:102077. [Crossref] [PubMed]
33. Dong Y, Qiu Y, Deng J, et al. Insufficient examined lymph node count underestimates staging in pN3a patients after curative gastrectomy: a multicenter study with external validation. J Cancer Res Clin Oncol 2020;146:515-28. [Crossref] [PubMed]
34. Huang JY, Xing YN, Wang X, et al. The Prognosis Value of Lymphatic Vessel Invasion in pN0 Gastric Cancer Patients with Insufficient Examined Lymph Nodes. J Gastrointest Surg 2020;24:299-306. [Crossref] [PubMed]
35. Smith DD, Schwarz RR, Schwarz RE. Impact of total lymph node count on staging and survival after gastrectomy for gastric cancer: data from a large US-population database. J Clin Oncol 2005;23:7114-24. [Crossref] [PubMed]
36. Volpe CM, Driscoll DL, Douglass HO Jr. Outcome of patients with proximal gastric cancer depends on extent of resection and number of resected lymph nodes. Ann Surg Oncol 2000;7:139-44. [Crossref] [PubMed]
37. Sano T, Coit DG, Kim HH, et al. Proposal of a new stage grouping of gastric cancer for TNM classification: International Gastric Cancer Association staging project. Gastric Cancer 2017;20:217-25. [Crossref] [PubMed]
38. Szczepanik AM, Paszko A, Szura M, et al. Alternative staging of regional lymph nodes in gastric cancer. Prz Gastroenterol 2016;11:145-9. [Crossref] [PubMed]
39. Zhou Y, Zhang J, Cao S, et al. The evaluation of metastatic lymph node ratio staging system in gastric cancer. Gastric Cancer 2013;16:309-17. [Crossref] [PubMed]
40. Zhu J, Xue Z, Zhang S, et al. Integrated analysis of the prognostic role of the lymph node ratio in node-positive gastric cancer: A meta-analysis. Int J Surg 2018;57:76-83. [Crossref] [PubMed]
41. Seeruttun SR, Xu L, Wang F, et al. A homogenized approach to classify advanced gastric cancer patients with limited and adequate number of pathologically examined lymph nodes. Cancer Commun (Lond) 2019;39:32. [Crossref] [PubMed]
42. Qiu MZ, Qiu HJ, Wang ZQ, et al. The tumor-log odds of positive lymph nodes-metastasis staging system, a promising new staging system for gastric cancer after D2 resection in China. PLoS One 2012;7:e31736. [Crossref] [PubMed]
43. Sun Z, Xu Y. Log odds of positive lymph nodes: a novel prognostic indicator superior to the number-based and the ratio-based N category for gastric cancer patients with R0 resection. Cancer 2010;116:2571-80. [Crossref] [PubMed]
44. Liu Y, Cui H, Song L, et al. The association of survival with adjuvant chemotherapy in patients with ypTNM stage I gastric cancer after neoadjuvant chemotherapy. World J Surg Oncol 2026;24:100. [Crossref] [PubMed]
45. He FJ, Zhang WH, Jin T, et al. Characteristics of Lymph Node Metastasis in Early Gastric Cancer: a Single-center Retrospective Study. Int J Surg 2025; Epub ahead of print. [Crossref] [PubMed]
46. Song P, Tian W, Zhao J, et al. Impact of Adjuvant Chemotherapy on Survival of the Patients with pT1N + M0 Gastric Cancer: A Retrospective Cohort Study. J Invest Surg 2026;39:2594559. [Crossref] [PubMed]
47. Wang XQ, Bao M, Zhang C. Prognostic value of different lymph node staging methods for node-positive cardia gastric cancer: a register-based retrospective cohort study. BMJ Open 2021;11:e050378. [Crossref] [PubMed]