Development of a postoperative risk scoring tool for elderly patients with early gastric cancer: an exploratory prognostic score derivation study

Introduction

Background

Gastric cancer remains the fifth most common malignancy and the fourth leading cause of cancer-related mortality worldwide, with approximately one million new cases diagnosed annually (1). In South Korea, a national endoscopic screening program has contributed to a high rate of early-stage detection, with early gastric cancer (EGC) now comprising the majority of newly diagnosed cases at high-volume centers (2). At such institutions, 5-year overall survival (OS) for EGC consistently exceeds 90% (3).

The proportion of elderly patients—broadly defined as aged ≥70 years—among all newly diagnosed gastric cancer cases has more than tripled over recent decades, rising from approximately 9.1% in 1995 to over 30% in 2019 (4). This trend reflects population aging and is expected to increase even more in East Asia and globally. Elderly patients represent a surgically heterogeneous population characterized by variable physiological reserve, comorbidity burden, and age-related functional decline, all of which can substantially influence surgical outcomes independently of tumor stage.

Rationale and knowledge gap

In advanced gastric cancer (AGC), prior studies have identified older age and total gastrectomy as key determinants of poor long-term survival after gastrectomy (5). However, evidence regarding surgical risk factors for long-term survival specifically in elderly EGC patients is notably limited. This knowledge gap is clinically consequential because for elderly EGC patients with high comorbidity burden, standard gastrectomy with D2 lymphadenectomy may carry postoperative risks disproportionate to its oncological benefit, particularly given the low probability of nodal metastasis in early-stage disease.

Several candidate variables—including the Charlson Comorbidity Index (CCI), Prognostic Nutritional Index (PNI), preoperative hemoglobin level, and extent of gastrectomy—have individually been associated with surgical outcomes in gastric cancer. However, their combined predictive utility for long-term survival in elderly EGC patients has not been systematically evaluated or formulated into a clinically actionable risk stratification tool (6,7).

Objective

This study aimed to: (I) systematically identify independent preoperative and surgical risk factors associated with 5-year OS in elderly patients (≥75 years) with stage I gastric cancer undergoing curative gastrectomy; and (II) derive and evaluate a simple composite risk scoring tool for perioperative risk stratification in this population. We present this article in accordance with the TRIPOD reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0470/rc).

Methods

Study design and patient eligibility

This was a single-center retrospective study conducted at Asan Medical Center, a high-volume tertiary center in South Korea. This study was designed as an exploratory derivation of a simplified prognostic risk score rather than as a validated prediction model. The study enrolled elderly EGC patients who received gastrectomy between January 2007 to December 2016. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The Institutional Review Board of Asan Medical Center approved the study protocol (No. 2022-1461) and waived the requirement for individual informed consent owing to the retrospective design.

Patients were enrolled if they: (I) were aged ≥75 years at the time of surgery; (II) had histopathologically confirmed stage I gastric adenocarcinoma following R0 resection; and (III) underwent curative-intent gastrectomy within the study period. Patients were excluded if they: (I) had prior gastric surgery; (II) received neoadjuvant chemotherapy or chemoradiotherapy; (III) underwent combined organ resection for direct tumor invasion; (IV) had a history of another primary malignancy within 5 years before surgery; or (V) underwent emergency operations. All of the patients were observed for post-operative complications and survival at 5 years.

Data collection and variable definitions

Clinical data were retrieved from the institutional prospectively maintained gastric cancer registry and electronic medical records. Baseline characteristics included age (dichotomized as 75–79 vs. ≥80 years), sex, body mass index (BMI, kg/m²), and American Society of Anesthesiologists (ASA) physical status classification. Comorbidity burden was quantified by the age-independent CCI (scored without the age component), which isolates chronic disease burden from chronological age; a high CCI was defined as score 3 or higher (8). Preoperative laboratory parameters, collected within four weeks of surgery, included serum albumin (g/dL), total lymphocyte count (/mm³), and hemoglobin (g/dL). The PNI was calculated as: 10 × serum albumin (g/dL) + 0.005 × total lymphocyte count (/mm³) (9). Anemia was defined per World Health Organization (WHO) sex-specific thresholds: hemoglobin <13 g/dL in males and <12 g/dL in females (10). Postoperative complications were graded using the Clavien–Dindo classification. Overall complication was defined as the occurrence of any complication (grade I–V).

Surgical variables included extent of gastrectomy (distal vs. total), operative approach (open vs. laparoscopic), extent of lymphadenectomy (limited vs. standard D2), and type of reconstruction. The extent of gastrectomy was defined as the preoperatively planned extent of resection. In gastric cancer, the planned procedure (distal or total gastrectomy) is determined before surgery on the basis of tumor location assessed by endoscopy and computed tomography. For example, tumors of the cardia or upper body are planned for total gastrectomy with near-complete certainty preoperatively. This variable was therefore treated as information available at the time of preoperative consultation. Pathological variables included tumor-node-metastasis (TNM) stage and lymph node metastasis status. Postoperative complications were recorded as any deviation from the expected postoperative course within 30 days of surgery, classified by the Clavien-Dindo system.

Outcome measures

The primary endpoint was 5-year OS, defined as the interval from the date of surgery to death from any cause or last known follow-up. Follow-up data were obtained from outpatient clinical records. Vital status was obtained from national health-insurance coverage records rather than from institutional surveillance visits. This source provides complete ascertainment of survival but does not record individual follow-up encounter dates; accordingly, individual follow-up duration and median follow-up time could not be calculated.

Risk score derivation

A composite risk score was constructed using the five variables independently identified as significant in the multivariable Cox analysis. Each factor was assigned a score of one point if present and zero if absent, yielding a total score ranging from zero to five.

Risk score = age ≥80 years (0 or 1) + male sex (0 or 1) + CCI ≥3 (0 or 1) + preoperative anemia (0 or 1) + total gastrectomy (0 or 1)

For ease of clinical application, each independent predictor was assigned an equal weight of one point; a coefficient-based weighting scheme was not used in order to preserve the simplicity of the score. Patients were classified into three groups based on clinically meaningful survival separation observed in Kaplan–Meier analysis: low-risk (score 0–1), intermediate-risk (score 2), and high-risk (score ≥3).

Statistical analysis

Continuous variables are presented as mean ± standard deviation (SD). Categorical variables are expressed as counts and proportions (%). Survival was analyzed using the Kaplan-Meier method; differences between groups were assessed by log-rank test. Cox proportional hazard regression was used for univariable and multivariable survival analysis; variables with P<0.10 on univariable analysis were entered into the multivariable model using backward elimination. Hazard ratio (HR) and 95% confidence interval (CI) are reported. A two-sided P<0.05 was considered statistically significant. All candidate predictors were derived from the mandatory preoperative work-up performed in every patient before gastrectomy, and vital status was obtained from national health-insurance records covering the entire population. The dataset was therefore complete, with no missing values for any predictor or for the survival outcome. All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and R version 4.2.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Patient and surgical characteristics

A total of 534 patients met eligibility criteria. Baseline clinical characteristics and survival outcomes are shown in Table 1. The mean age was 77.6±2.8 years; 107 patients (20.1%) were aged 80 years or older. The cohort comprised 366 males (68.5%) and 168 females (31.5%). Most patients had ASA class 2 (89.3%). The age-independent CCI was 3 or higher in 155 patients (29.0%). The most common comorbidities were diabetes mellitus (17.6%) and cerebrovascular disease (5.8%). A total of 373 patients (69.9%) had at least one comorbidity. Preoperative anemia was present in 213 patients (39.9%). There were a total of 6 (1.1%) recurrences and 92 (17.2%) deaths within the 5-year observation period.

Surgical and pathological characteristics are presented in Table 2. A laparoscopic approach was used in 299 patients (56.0%). Distal gastrectomy was performed in 450 patients (84.3%) and total gastrectomy in 84 (15.7%). The majority of patients had stage IA disease (76.8%) and 90.4% had no lymph node metastasis (pN0), with the mean number of harvested lymph nodes being 29.4±11.9. Postoperative complications occurred in 129 patients (24.2%). There was no 30-day mortality, however one patient (0.2%) expired within ninety days of surgery.

Univariable and multivariable Cox regression analysis

Cox regression results are presented in Table 3. On univariable analysis, factors significantly associated with reduced OS included age ≥80 years (HR 2.2, 95% CI: 1.4–3.4; P<0.001), male sex (HR 2.5, 95% CI: 1.4–4.4; P=0.001), CCI ≥3 (HR 1.6, 95% CI: 1.1–2.5; P=0.02), preoperative anemia (HR 2.0, 95% CI: 1.3–3.0; P=0.001), total gastrectomy (HR 2.2, 95% CI: 1.4–3.4; P=0.001), and postoperative complications (HR 2.0, 95% CI: 1.3–3.1; P=0.001). BMI, PNI, diabetes mellitus, previous abdominal surgery, extent of lymphadenectomy, and TNM stage were not significantly associated with OS.

On multivariable Cox regression, five variables independently predicted reduced 5-year OS: age ≥80 years (HR 2.0, 95% CI: 1.3–3.1; P=0.002), male sex (HR 2.1, 95% CI: 1.2–3.7; P=0.009), age-independent CCI ≥3 (HR 1.6, 95% CI: 1.1–2.4; P=0.03), preoperative anemia (HR 1.7, 95% CI: 1.1–2.6; P=0.01), and total gastrectomy (HR 2.0, 95% CI: 1.3–3.2; P=0.003). Postoperative complications did not retain independent significance after adjustment. Kaplan-Meier survival curves for each independent risk factor are shown in Figure 1.

Figure 1 Kaplan-Meier overall survival curves for each of the five independent prognostic risk factors identified by multivariable Cox regression analysis. (A) Age: 75–79 vs. ≥80 years (HR 2.0, P<0.001); (B) sex: female vs. male (HR 2.1, P<0.001); (C) age-independent CCI: score 2 vs. ≥3 (HR 1.6, P=0.02); (D) hemoglobin: normal (≥13 g/dL for male, ≥12 g/dL for female) vs. anemia (<13 g/dL for male, <12 g/dL for female) (HR 1.7, P=0.001); (E) extent of gastrectomy: distal vs. total (HR 2.0, P<0.001). P values by log-rank test. CCI, Charlson Comorbidity Index; Hb, hemoglobin; HR, hazard ratio.

Risk score stratification and survival outcomes

Based on the five independent variables, a composite risk score was assigned (1 point each; range 0–5). Score distribution: 252 patients (47.2%) were in the low-risk group (score 0–1), 184 (34.5%) in the intermediate-risk group (score 2), and 98 (18.4%) in the high-risk group (score ≥3). Table 4 presents Kaplan-Meier OS estimates at 1, 3, and 5 years by risk category. The overall 5-year OS for the entire cohort was 82.7%. 5-year OS was 91.2% (low risk), 82.5% (intermediate risk), and 61.2% (high risk), with a highly significant difference across strata (log-rank P<0.0001). Kaplan-Meier survival curves stratified by risk score category are shown in Figure 2.

Figure 2 Kaplan-Meier overall survival curves stratified by composite postoperative risk score category. Low risk (score 0–1, n=252; red line), intermediate risk (score 2, n=184; green line), and high risk (score ≥3, n=98; blue line). 5-year overall survival rates: 91.2%, 82.5%, and 61.2%, respectively (log-rank P<0.0001). Risk score = age ≥80 years + male sex + CCI ≥3 + preoperative anemia + total gastrectomy (1 point each; total range 0–5). CCI, Charlson Comorbidity Index.

Discussion

Key findings

This retrospective cohort study of 534 elderly patients with stage I gastric cancer presents the first composite risk scoring tool for stratifying long-term survival in elderly EGC patients after curative gastrectomy. Five independent variables—age ≥80 years, male sex, high age-independent CCI (≥3), preoperative anemia, and total gastrectomy—each associated with approximately a 2-fold increase in mortality hazard, were combined into a simple five-point score. This score distinguished three groups with 5-year OS rates of 91%, 83%, and 61% for low-, intermediate-, and high-risk patients, respectively, with a significant difference (P<0.0001).

Strengths and limitations

The principal strength of this study is its large, well-characterized cohort from a high-volume tertiary center with a prospectively maintained surgical registry, lending confidence to the data quality. The deliberate use of the age-independent CCI—rather than the standard age-adjusted CCI—is a key methodological strength, allowing independent separation of chronological age and chronic disease burden as prognostic constructs, both of which independently entered the multivariable model (8). All five score variables are preoperatively measurable without specialized examinations, making the tool easy to use in clinical practice.

This study has several limitations. First, as a retrospective single-center study, selection bias may be present and findings may not generalize to institutions with different patient profiles or surgical volumes. Second, because the score was derived and evaluated within the same single-center cohort, our results are susceptible to overfitting and optimism. We therefore present this score as an exploratory, hypothesis-generating tool that requires internal validation, formal assessment of discrimination and calibration, and external validation in independent cohorts before it can be applied in clinical practice. Third, formal frailty and geriatric assessments were not evaluated. Validated measures such as Eastern Cooperative Oncology Group performance status, the Clinical Frailty Scale, sarcopenia, activities of daily living, and comprehensive geriatric assessment are strongly associated with outcomes in elderly surgical patients and may outperform chronological age; these were not routinely recorded during the study period and could not be incorporated. The proposed score should therefore be regarded as complementary to, rather than a replacement for, comprehensive geriatric assessment, and its performance relative to established frailty metrics remains to be determined. Fourth, each predictor was weighted equally despite differing hazard ratios. We did not formally compare the simplified equally weighted score with a coefficient-weighted model, and this simplification may have reduced discriminative performance. Future work should evaluate whether coefficient-based weighting meaningfully improves prediction. Last but not least, as the exact cause of death data were not available—vital status was obtained from national insurance records—we were unable to perform a competing-risk analysis.

Comparison with similar research

Previous studies examining prognostic factors in elderly gastric cancer patients have predominantly focused on AGC or mixed-stage populations (5,11). In AGC, older age, total gastrectomy, and high comorbidity burden have been consistently identified as predictors of poor surgical outcomes, consistent with the present findings. However, the direct translation of these findings to the EGC setting is limited by the fundamentally different disease trajectory and operative rationale. In the oncological context of EGC, where excellent OS is expected in younger patients, the pronounced survival differential identified in our high-risk group (61.2% at 5 years) underscores the extent to which host factors can dominate prognosis in elderly patients with otherwise-favorable tumor characteristics.

The present risk score conceptually aligns with preoperative risk stratification tools used in other surgical oncology settings—such as the Lee Cardiac Risk Index or the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) risk calculator—in its aim to aggregate multiple clinical risk factors into a single actionable score (12). Its novelty lies in its specific derivation and application in elderly EGC, a population that is insufficiently addressed.

Explanations of findings

The independent prognostic significance of age 80 years and older within an already-elderly cohort reflects the clinically meaningful physiological differences between patients aged 75 to 79 years and octogenarians. Octogenarians exhibit substantially diminished functional reserve and an accelerated trajectory of age-related decline, rendering them a distinct subgroup for surgical risk assessment (13). The age-independent CCI independently contributed prognostic information beyond age itself, confirming that chronic disease burden—particularly cardiovascular, renal, and metabolic comorbidities—exerts a distinct and additive adverse effect on long-term survival (8).

Preoperative anemia, the sole modifiable risk factor, is associated with impaired tissue oxygenation, reduced functional reserve, compromised wound healing, and higher vulnerability to perioperative complications (14). In elderly patients, anemia often reflects broader nutritional compromise or incipient frailty, making it both a prognostic marker and a potential therapeutic target. Correction of anemia through preoperative iron supplementation, erythropoiesis-stimulating agents, or transfusion may attenuate perioperative risk and deserves prospective evaluation in this population.

Total gastrectomy was associated with a two-fold higher mortality hazard compared to distal gastrectomy after multivariable adjustment, consistent with its greater physiological burden, including higher rates of nutritional deficiency, malabsorption, weight loss, and early postoperative morbidity (15). Where anatomically feasible, distal gastrectomy should be preferred in elderly EGC patients.

Implications and actions needed

The clinical utility of this risk score lies in enabling objective, structured preoperative risk communication and personalized surgical planning. Patients scoring 0–1 are at low risk and standard curative gastrectomy is appropriate. Patients scoring 3 or higher face substantially elevated long-term mortality risk, warranting a careful multidisciplinary discussion that considers less invasive alternatives, geriatric assessment, and detailed informed consent.

For high-risk elderly EGC patients, stomach-preserving surgical approaches may offer a more favorable risk-benefit profile. For example, laparoscopic and endoscopic cooperative surgery (LECS) enables full-thickness local resection under combined laparoscopic and endoscopic guidance, avoiding extensive lymphadenectomy (16). Also, laparoscopic sentinel node navigation surgery (LSNNS), which uses dual-tracer mapping with indocyanine green and Tc-99m radiocolloid, facilitates tailored lymph node dissection and stomach preservation (17). The multicenter randomized SENORITA trial demonstrated comparable survival outcomes with LSNNS versus standard gastrectomy in selected EGC patients (18), though data specific to elderly populations are absent. Comparative effectiveness studies of standard versus stomach-preserving surgery in high-risk elderly EGC patients are a critical and currently unmet research priority.

Taking these results into account, preoperative hemoglobin should be systematically evaluated and corrected if possible in all elderly EGC patients preparing for surgery. Prospective multicenter studies validating this score and comparing surgical strategies in high-risk elderly EGC patients are warranted.

Conclusions

Five negative prognostic risk factors were identified for elderly EGC patients undergoing curative gastrectomy: age ≥80 years, male sex, age-independent CCI ≥3, preoperative anemia, and total gastrectomy. A composite risk score derived from these factors stratified 5-year OS into low (91.2%), intermediate (82.5%), and high-risk (61.2%) groups (P<0.0001). This practical tool may support personalized surgical decision-making and patient counseling in elderly EGC patients.

Acknowledgments

The abstract of this paper was published in the KINGCA (The Korean International Gastric Cancer Week) 2025 conference.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0470/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0470/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0470/prf

Funding: This study was supported by the National R&D Program for Cancer Control through the National Cancer Center (NCC) funded by the Ministry of Health & Welfare, Republic of Korea (No. RS-2025-02263269).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-0470/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 Institutional Review Board of Asan Medical Center (No. 2022-1461), and individual informed consent was waived due to the retrospective study design.

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. Park EH, Jung KW, Park NJ, et al. Cancer Statistics in Korea: Incidence, Mortality, Survival, and Prevalence in 2022. Cancer Res Treat 2025;57:312-30. [Crossref] [PubMed]
3. Japanese Gastric Cancer Association. Japanese Gastric Cancer Treatment Guidelines 2021 (6th edition). Gastric Cancer 2023;26:1-25.
4. Liu W, Peng ZZ, Zhang T, et al. Burden, trends, driving factors, and predictions of gastric cancer: a cross-national comparative analysis of China, Japan, and South Korea. BMC Cancer 2025;25:1669. [Crossref] [PubMed]
5. Endo S, Higashida M, Furuya K, et al. Prognostic factors for gastric cancer patients aged ≥ 85 years. BMC Cancer 2024;24:745. [Crossref] [PubMed]
6. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 2011;173:676-82. [Crossref] [PubMed]
7. Deng H, He Y, Huang G, et al. Predictive value of prognostic nutritional index in patients undergoing gastrectomy for gastric cancer: A systematic review and meta-analysis. Medicine (Baltimore) 2024;103:e39917. [Crossref] [PubMed]
8. Maezawa Y, Aoyama T, Kano K, et al. Impact of the Age-adjusted Charlson comorbidity index on the short- and long-term outcomes of patients undergoing curative gastrectomy for gastric cancer. J Cancer 2019;10:5527-35. [Crossref] [PubMed]
9. Feng JF, Chen QX. Significance of the prognostic nutritional index in patients with esophageal squamous cell carcinoma. Ther Clin Risk Manag 2014;10:1-7. [Crossref] [PubMed]
10. World Health Organization. Nutritional anemias: tools for effective prevention and control. Geneva: WHO; 2017.
11. Paredero-Pérez I, Jimenez-Fonseca P, Cano JM, et al. State of the scientific evidence and recommendations for the management of older patients with gastric cancer. J Geriatr Oncol 2024;15:101657. [Crossref] [PubMed]
12. Bilimoria KY, Liu Y, Paruch JL, et al. Development and evaluation of the universal ACS NSQIP surgical risk calculator: a decision aid and informed consent tool for patients and surgeons. J Am Coll Surg 2013;217:833-42.e1-3.
13. Loh KP, Liposits G, Arora SP, et al. Adequate assessment yields appropriate care-the role of geriatric assessment and management in older adults with cancer: a position paper from the ESMO/SIOG Cancer in the Elderly Working Group. ESMO Open 2024;9:103657. [Crossref] [PubMed]
14. Aoyama T, Hashimoto I, Maezawa Y, et al. Perioperative Anemia Is an Independent Prognostic Factor for Gastric Cancer Patients Who Receive Curative Treatment. Anticancer Res 2024;44:5551-7. [Crossref] [PubMed]
15. Calì M, Bona D, De Bernardi S, et al. Impact of Anastomotic Leak on Long-Term Survival After Gastrectomy: Results from an Individual Patient Data Meta-Analysis. Cancers (Basel) 2025;17:2471. [Crossref] [PubMed]
16. Hiki N, Nunobe S, Matsuda T, et al. Laparoscopic endoscopic cooperative surgery. Dig Endosc 2015;27:197-204. [Crossref] [PubMed]
17. Mourdi N, Wu Y, Su Y, et al. The role of indocyanine green in the intraoperative navigation of gastric cancer surgery: a systematic review and meta-analysis. BMC Cancer 2025;25:1920. [Crossref] [PubMed]
18. Hur H, Lee YJ, Kim YW, et al. Clinical Efficacy of Laparoscopic Sentinel Node Navigation Surgery for Stomach Preservation in Patients With Early Gastric Cancer: 5-year Results of the SENORITA Trial. Ann Surg 2024; Epub ahead of print. [Crossref]