Abstract
Background/Aim: Colorectal cancer (CRC) presents a significant challenge in oldest-old patients (≥85 years), where surgical intervention carries substantial perioperative risks. Nutritional status is a crucial determinant of outcomes, and the Geriatric Nutritional Risk Index (GNRI) has shown promise. This prospective study aimed to validate the GNRI as a key indicator of perioperative outcomes in oldest-old patients undergoing CRC surgery, and to establish its utility in preoperative risk stratification.
Patients and Methods: This prospective study enrolled patients aged ≥85 years undergoing elective surgery for CRC. Preoperative GNRI was calculated using the formula: GNRI=14.89×serum albumin (g/dl)+41.7×[actual body weight/ideal body weight (corresponding to body mass index 22)]. Patients were stratified into two groups: GNRI >98 and GNRI ≤98. Baseline demographics, clinical characteristics, geriatric assessments (including Geriatric-8 and EuroQol 5 dimension), and postoperative complication rates were analyzed.
Results: Twenty-four patients (median age 88 years, interquartile range=86-91) were included: 11 in the GNRI >98 group and 13 in the GNRI ≤98 group. The patients with GNRI >98 demonstrated significantly better G8 scores (median 12 vs. 11, p<0.01) and EQ-5D index values (median 88 vs. 75.0, p<0.01). The postoperative complication rate was significantly higher in the GNRI ≤98 group (p=0.02).
Conclusion: Preoperative GNRI effectively identifies oldest-old patients with CRC at increased risk for postoperative complications. A GNRI ≤98 correlates with poorer nutritional status and impaired geriatric functional parameters. These findings highlight GNRI’s utility as a simple, valuable tool for preoperative risk stratification, potentially guiding interventions to optimize outcomes in this vulnerable population.
Introduction
Colorectal cancer (CRC) represents a significant global health challenge, particularly affecting the older population (1). While surgical resection remains the primary curative treatment for localized CRC, advanced age is associated with increased postoperative morbidity and mortality (2). Although laparoscopic techniques have improved surgical outcomes in older patients (3-5), accurate preoperative risk assessment remains crucial for optimal patient care.
Nutritional status has emerged as a critical determinant of surgical outcomes, with the geriatric nutritional risk index (GNRI) showing particular promise as a risk stratification tool (6-8). The GNRI, which incorporates serum albumin levels and body weight, offers a simple, cost-effective method for assessing nutritional risk in older individuals (9). Recent studies have demonstrated its significant prognostic value in older CRC patients undergoing surgery, with low preoperative GNRI correlating with increased postoperative complications and decreased long-term survival (6, 7, 10, 11).
The clinical utility of GNRI extends beyond surgical outcomes, showing predictive value in various therapeutic contexts, including response to adjuvant chemotherapy and neoadjuvant chemoradiotherapy (12, 13). Meta-analyses have consistently validated GNRI’s predictive capabilities in patients with CRC (14). Furthermore, recent research has established GNRI’s prognostic significance in specific clinical scenarios, such as predicting outcomes following stent placement for obstructive CRC and survival in metastatic disease (15).
Despite this growing body of evidence supporting GNRI’s clinical relevance, prospective studies specifically examining its role in older patients with CRC remain limited. This prospective study aimed to validate GNRI as a key indicator of surgical outcomes in older adults undergoing CRC surgery. By examining short-term outcomes, we aimed to establish GNRI’s utility in preoperative risk stratification and its potential role in guiding nutritional interventions to enhance patient outcomes.
Patients and Methods
Patients. This was a single-center, prospective, observational study conducted in the perioperative management center (PERiO), Okayama University Hospital. We followed the recommendations of the Equator Network and complied with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Colorectal cancer surgery patients over 85 years of age who were treated at PERiO between July 1, 2020, and March 31, 2024, excluding emergent surgery cases, were included. The following information was obtained from routine medical care. Basic patient characteristics, surgical findings, and postoperative outcomes were collected from electronic medical records.
Ethical approval. All procedures performed were in accordance with the ethical standards of our institutional research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. This study was approved by our Institutional Review Board (approval No. 2006-040). Informed consent was obtained from all study participants.
Nutritional evaluation, geriatric assessment, and outcome measurement. The data of comprehensive geriatric assessments were collected by PERiO nutritionists and nurses using the geriatric nutritional risk index (GNRI) and geriatric 8 (G8) during the preoperative interview. Patients were also asked to complete a self-administered questionnaire containing items regarding their living status, Japanese version of the EuroQol 5 dimension 5-level (EQ-5D-5L), and Eastern Cooperative Oncology Group Performance Status (ECOG-PS). All patients included in the study underwent consultation with a PERiO-affiliated nurse prior to receiving treatment at our facility as previously described (16). Baseline GNRI was calculated from serum albumin and body mass index (BMI) obtained on hospital admission as previously described (17); GNRI= 14.89× serum albumin (g/dl)+41.7×BMI/22. GNRI was categorized into two nutritional statuses: GNRI >98 (normal risk); ≤98 (nutritional risk). The G8 is a screening tool to assess physical function, medications, nutrition, and mood in older people and is considered most useful because of its high sensitivity and acceptable specificity (18). EQ-5D-5L (Japanese version) is a five-item, five-choice questionnaire of the patient’s health status. The questionnaire is self-administered, requires approximately 5 min to complete, and provides a numerical evaluation of quality of life (QoL) (19). The primary outcome was the incidence of all postoperative complications according to the Clavien–Dindo classification (20). Complications were recorded for 30 days after surgery by the physician-in-charge. The secondary outcome was length of postoperative hospital stay.
Statistical analysis. Continuous variables are described using medians and interquartile ranges (IQRs), and categorical variables are described using counts and percentages. Descriptive statistics were obtained for the incidence of postoperative complications, length of postoperative hospital stay, and discharge home rate. Correlations between patient information and observational and measurement items were also analyzed. All statistical analyses were conducted using GraphPad Prism 6 (GraphPad Software, Boston, MA, USA). p<0.05 was considered statistically significant. Because this study has an exploratory component, a multiplicity of tests was not considered.
Results
Baseline characteristics. Table I presents the baseline characteristics of the 24 oldest-old patients (≥85 years) who underwent surgery for colorectal cancer. The median age of the entire cohort was 89 years [interquartile range (IQR)=86-91], with 10 (41.7%) males and 14 (58.3%) females. Patients were stratified based on their GNR into two groups: GNRI >98 (n=11) and GNRI ≤98 (n=13). No statistically significant differences were observed between the two GNRI groups with respect to age (p=0.21) and sex (p =0.24). The median body mass index (BMI) was significantly higher in the GNRI >98 group [23.3 kg/m2 (IQR=21.3-24.9)] compared to the GNRI ≤98 group [19.1 kg/m2 (IQR=17.5-22.6); p<0.01]. The history of prior abdominal surgery and tumor location (colon vs. rectum) were comparable between the groups (p=0.81 and p=0.23, respectively). A statistically significant difference was found in the median serum albumin levels, with the GNRI >98 group exhibiting higher levels [4.1 g/dl (IQR=4.0-4.3)] compared to the GNRI ≤98 group [3.6 g/dl (IQR=3.2-3.9); p<0.01]. The distribution of pathological stage (pStage I-III) was not significantly different between the two GNRI groups (p=0.44).
Baseline characteristics of 24 patients undergoing minimum invasive surgery for colorectal cancer.
Geriatric assessment. Table II details further baseline geriatric assessments of the study cohort. The median G8 score was significantly higher in the GNRI >98 group [12 (IQR=11.0-14.0)] compared to the GNRI ≤98 group [(IQR=8.8-11.8); p=0.01]. In contrast, no significant difference was observed in the median Instrumental Activities of Daily Living (IADL) score between the two groups (p=0.97). The median EQ-5D index, a measure of health-related quality of life, was significantly higher in the GNRI >98 group [88 (IQR=74-93)] compared to the GNRI ≤98 group [75.0 (IQR=50-80); p<0.01]. The proportion of patients living alone did not differ significantly between the groups (p=0.23).
Baseline geriatric assessment.
The ECOG-PS was also comparable between the groups (p=0.19), with the majority of patients in both the GNRI >98 (90.9%) and GNRI ≤98 (69.2%) groups having an ECOG-PS of <2.
Perioperative features. Table III presents an overview of the perioperative interventions and immediate outcomes. A significantly higher proportion of patients in the GNRI ≤98 group (84.6%) underwent prehabilitation intervention compared to the GNRI >98 group (45.5%; p=0.04). Similarly, preoperative oral care by the perioperative dental support team was more frequently performed in the GNRI >98 group (45.5%) than in the GNRI ≤ 98 group (7.7%; p=0.03). The choice of surgical approach (laparoscopy vs. robot-assisted surgery) did not differ significantly between the two groups (p=0.47). Laparoscopic surgery was the predominant approach in both the GNRI >98 (72.3%) and GNRI ≤98 (84.6%) groups. No statistically significant differences were observed between the groups in terms of median operation time (p=0.70) or median estimated blood loss (p=0.61). Furthermore, the rate of intensive care unit (ICU) admission postoperatively was also comparable between the GNRI >98 (27.3%) and GNRI ≤98 (46.2%) groups (p=0.34).
Associations of GNRI status and clinicopathological factors in oldest-old patients with colorectal cancer.
Correlation between GNRI and short-term outcomes. Table IV details the postoperative complications and outcomes. A statistically significant difference was observed in the incidence of Clavien-Dindo (C-D) Grade I-V complications, with a higher proportion in the GNRI ≤98 group (38.5%) compared to the GNRI >98 group (0%; p=0.02). Specifically, postoperative delirium and intra-pelvic abscesses were exclusively observed in the GNRI ≤98 group (15.4% each). Ileus and hyponatremia each occurred in one patient in the GNRI ≤98 group (7.7%). No C-D Grade I-V complications occurred in the GNRI >98 group. The incidence of more severe complications (C-D Grade ≥II) was low in both groups, with one patient in each group experiencing such a complication (p=0.34). The median postoperative hospital stay was 11 days in the entire cohort (IQR=9-14) and did not differ significantly between the GNRI >98 [9 days (IQR=8-14)] and GNRI ≤98 [11 days (IQR=9-13)] groups (p=0.51). Similarly, the rate of direct discharge to home was high in both groups, with no significant difference observed (GNRI >98: 100%; GNRI ≤98: 84.6%; p=0.17). Notably, there were no readmissions within 30 days in either group.
Correlations of GNRI status and postoperative outcomes in oldest-old patients with colorectal cancer.
Discussion
Our study investigated the utility of the GNRI as a predictor of perioperative outcomes in oldest-old patients (≥85 years) undergoing surgery for colorectal cancer. The principal findings revealed that a lower preoperative GNRI (≤98) was significantly associated with a higher incidence of postoperative complications, particularly Clavien-Dindo Grade I-V complications (38.5% vs. 0%, p=0.02). Specifically, episodes of postoperative delirium and intra-pelvic abscesses were exclusively observed in the lower GNRI group. Furthermore, patients with a lower GNRI exhibited a significantly poorer scores in the G8 geriatric assessment and EQ-5D health-related QoL index at baseline. These results underscore the potential of the GNRI as a readily available and valuable tool for identifying oldest-old patients with colorectal cancer at increased risk of adverse postoperative events.
The significance of our findings lies in highlighting the crucial role of nutritional status in the surgical outcomes of this vulnerable population. As the number of oldest-old patients undergoing major surgical procedures continues to rise, accurate risk stratification becomes paramount for optimizing patient care and resource allocation. Our study adds to the growing body of evidence, as cited, supporting the GNRI as a key indicator in this context.
Previous reports have demonstrated an association between lower preoperative GNRI and increased postoperative complications in older patients with colorectal cancer (6-8). Our results align with these findings, extending them specifically to the oldest-old subgroup and emphasizing the early postoperative period. Identifying patients with a low GNRI preoperatively allows for timely implementation of targeted interventions, such as prehabilitation programs and nutritional support, which may potentially mitigate the risk of postoperative complications and improve overall outcomes. In the present study, while the prehabilitation intervention rate was significantly higher in cases with lower GNRI, dental interventions were significantly less frequent. Although the underlying cause for the reduced dental interventions remains unclear, periodontal disease has been reported as a risk factor for infectious complications following gastrointestinal cancer surgery (21), and perioperative oral care is crucial for preventing the development of postoperative complications (22).
While our findings suggest a strong association between lower GNRI and increased postoperative morbidity, alternative interpretations warrant consideration. It is possible that the GNRI, while indicative of nutritional status, may also be a surrogate marker for other underlying frailty components or comorbidities not fully captured in our baseline assessments. For instance, subclinical inflammation or subtle impairments in physiological reserve, which are more prevalent in malnourished individuals, could independently contribute to poorer postoperative outcomes. Although we included measures like the G8 score and ECOG-PS, residual confounding by unmeasured factors cannot be entirely excluded. Further research incorporating more comprehensive frailty assessments and inflammatory markers could help delineate the specific contribution of nutritional risk, as assessed by GNRI, to postoperative complications in this population.
Study limitations. First, the sample size of 24 patients, while providing statistically significant results for key outcomes, is relatively small and from a single institution, potentially limiting the generalizability of our findings to broader populations and different healthcare settings. Future multi-center studies with larger cohorts are needed to validate these results. Second, our study focused primarily on short-term postoperative outcomes. While the incidence of early complications is critical, long-term outcomes such as recurrence-free survival and overall survival were not assessed. While lower GNRI has been reported to correlate with poor prognosis in elderly patients with colorectal cancer aged 65 years or older and 75 years or older (6-14, 23, 24), there are no reports on the prognostic value of GNRI specifically in the oldest-old population aged 85 years and above, necessitating longitudinal studies. Third, the GNRI, as acknowledged in previous literature, relies solely on serum albumin levels and body weight measurements, potentially overlooking other crucial aspects of nutritional status, such as micronutrient deficiencies and muscle mass. Comprehensive nutritional assessments, including these parameters, could provide a more holistic evaluation of a patient’s nutritional risk. Finally, the prehabilitation interventions were more frequently implemented in the lower GNRI group, which, while intended to improve outcomes in this higher-risk population, could potentially have influenced the observed complication rates. However, despite this, the lower GNRI group still experienced a significantly higher rate of complications.
Despite these limitations, this study provides valuable insight into the utility of GNRI as a predictor of short-term postoperative outcomes in oldest-old patients with colorectal cancer. Future research should focus on validating our findings in larger, multi-institutional cohorts and exploring the potential benefits of targeted preoperative interventions based on GNRI assessment. Investigating the role of more comprehensive nutritional assessments in refining risk stratification is also warranted. Furthermore, exploring the association between GNRI and long-term oncological outcomes in this population could provide a more complete understanding of its clinical significance. In clinical practice, the GNRI can serve as a simple and cost-effective tool to identify high-risk oldest-old patients who may benefit from intensified perioperative care, including optimized nutritional support and tailored prehabilitation strategies, ultimately aiming to improve surgical outcomes in this challenging patient population.
Conclusion
Our study demonstrates that a lower preoperative GNRI is significantly associated with increased postoperative complications in oldest-old patients with colorectal cancer. The GNRI serves as a readily available tool for identifying high-risk individuals who may benefit from tailored perioperative strategies to improve surgical outcomes in this vulnerable population. Further large-scale studies are warranted to validate these findings and explore the impact of targeted interventions.
Acknowledgements
The Authors would like to thank all the patients and collaborating physicians.
Footnotes
Authors’ Contributions
Conception and Design: F. Teraishi, M. Utsumi, Y. Yoshida, R. Shoji, N. Kanaya, Y. Matsumi, K. Shigeyasu, Y. Kondo, S. Itagaki, R. Tamura, Y. Matsuoka. Data Collection: F. Teraishi, S. Itagaki, R. Tamura. Analysis and Interpretation of data: F. Teraishi. Manuscript Writing: F. Teraishi. Supervision: M. Inagaki, T. Fujiwara. All Authors have approved the final article.
Conflicts of Interest
All Authors declare that they have no conflicts of interest in relation to this study.
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (ChatGPT, OpenAI) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received June 7, 2025.
- Revision received June 20, 2025.
- Accepted June 23, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
