Published online Jan 14, 2027. doi: 10.3748/wjg.123195
Revised: June 27, 2026
Accepted: August 14, 2026
Published online: January 14, 2027
Processing time: 184 Days and 2.6 Hours
Systemic inflammation and nutritional status are closely related to tumor progres
To evaluate the associations of SINRIs with post-operative complications and OS in CRC and develop a machine learning (ML) prediction model.
This multicenter cohort study included 10205 patients who underwent radical surgery for CRC at four tertiary hospitals in Southwest China between 2011 and 2024. Associations of 14 SINRIs with post-operative complications and OS were assessed using multivariable logistic and Cox regression models, while metabolic dysfunction indicators were evaluated in a sub-cohort analysis. Restricted cubic spline analyses were performed to assess dose-response relationships. Six ML models were compared for predictive performance using the Harrell’s concordance index (C-index) and interpreted using SHapley Additive exPlanations.
Most SINRIs were significantly associated with post-operative complications and OS. Higher levels of inflammatory indicators were associated with increased risks of complications and mortality, whereas protective markers such as lymphocyte-to-monocyte ratio, prognostic nutritional index, and hemoglobin, albumin, lymphocyte, and platelet score were associated with better outcomes. Exploratory sub-cohort analysis suggested that triglyceride-glucose (TyG), TyG-body mass index, and metabolic score for insulin resistance were associated with worse OS. Among six candidate models, gradient boosting survival analysis (GBSA) showed the most stable predictive performance, with C-index values of 0.808 in the training cohort and 0.739 and 0.723 in the two external validation cohorts.
SINRIs are closely associated with post-operative complications and OS in CRC, and the GBSA model demon
Core Tip: This multicenter cohort study included 10205 patients with colorectal cancer (CRC) who underwent radical surgery. We systematically evaluated systemic inflammation- and nutrition-related indicators in relation to postoperative complications and overall survival, and further assessed metabolic dysfunction indices in a subcohort. Machine learning models were developed and externally validated, with gradient boosting survival analysis showing stable performance across centers. These findings support the integration of inflammatory, nutritional, and metabolic indicators into individualized risk stratification for CRC.
- Citation: Wang L, Liu C, Xie J, Liu JS, Zhou X, Liu F, Xiang CY, Zhao JN, Liu XR, Liu XY, Kang B, Cheng Y, Peng D. Integrating immune-inflammatory and nutritional indicators into machine learning-driven risk stratification for colorectal cancer. World J Gastroenterol 2027; 33(2): 123195
- URL: https://www.wjgnet.com/1007-9327/full/v33/i2/123195.htm
- DOI: https://dx.doi.org/10.3748/wjg.123195
Colorectal cancer (CRC) remains one of the most common malignancies worldwide and a leading cause of cancer-related morbidity and mortality[1]. Despite advances in surgical techniques and oncological therapies, many patients still experience post-operative complications and unfavorable long-term survival outcomes[2,3]. Therefore, accurate risk stratification is essential for optimizing perioperative management, follow-up, and individualized treatment decisions.
Systemic inflammation and nutritional status have been increasingly recognized as important determinants of cancer progression and prognosis. Chronic inflammation promotes tumor initiation and progression through genomic instabi
Metabolic dysfunction has also emerged as an important factor influencing CRC outcomes. Insulin resistance and dysr
Moreover, conventional regression models may be limited in capturing complex interactions and nonlinear relation
Therefore, in this multicenter cohort study, we evaluated the associations of SINRIs with post-operative complications and overall survival (OS) in patients with CRC. We also conducted an exploratory subcohort analysis to assess the prognostic role of metabolic dysfunction indicators. Finally, we developed and externally validated ML-based models integrating SINRIs and clinicopathological variables for individualized survival prediction (Figure 1).
This multicenter cohort study enrolled 10205 patients with CRC from four clinical centers in Southwest China between 2011 and 2024, including the First Affiliated Hospital of Chongqing Medical University (center A), the Second Affiliated Hospital of Chongqing Medical University (center B), the Affiliated Yongchuan Hospital of Chongqing Medical University (center C), and Chongqing Hospital of Jiangsu Province Hospital (center D). The inclusion criteria were as follows: (1) CRC diagnosed by histopathological examination; (2) Age ≥ 18 years; and (3) CRC as a single primary lesion. The exclusion criteria were as follows: (1) History of previous malignancy; and (2) Did not receive surgical treatment or only received palliative surgery. The flowchart of patient selection is shown in Figure 2. The study adhered to the ethical principles of the Declaration of Helsinki and was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital of Chongqing Medical University (Approval No. ZZ2026-0305-01). Since this study was a retrospective cohort study and the data used were de-identified and anonymized, the Ethics Committee approved a waiver of informed consent.
Clinical information included demographic characteristics, laboratory parameters, tumor characteristics, and outcomes. Demographic information included age, sex, smoking status, alcohol consumption, history of type 2 diabetes mellitus (T2DM), hypertension, coronary heart disease (CHD), and history of neoadjuvant therapy (NAT). Laboratory examination data were obtained from routine preoperative laboratory tests performed within 3 days before surgery, including complete blood counts, serum biochemical profiles, coagulation function tests, and tumor marker measurements. Tumor stage was determined according to the American Joint Committee on Cancer/Union for International Cancer Control tumor node metastasis staging system[20]. For outcome variables, overall complications were defined as grade ≥ II, and major complications were defined as grade ≥ III according to the Clavien-Dindo classification[21]. In addition, post-operative complications were categorized by organ system (respiratory, cardiovascular, digestive, and nutritional and metabolic). Detailed definitions are provided in the Supplementary material. Follow-up information included patient survival status and survival time. OS was defined as the time from the date of surgery to death from any cause or the date of the last follow-up. The last follow-up date was October 1, 2025. Follow-up data were primarily obtained through outpatient visits or telephone interviews.
Fourteen SINRIs, including NLR, SII, PNI, ALBI, platelet-to-lymphocyte ratio (PLR), lymphocyte-to-monocyte ratio (LMR), systemic inflammation response index (SIRI), pan-immune-inflammation value (PIV), hemoglobin, albumin, lymphocyte, and platelet score (HALP), fibrinogen-to-albumin ratio (FAR), platelet-to-albumin ratio (PAR), neutrophil-to-platelet ratio (NPR), albumin-to-globulin ratio (AGR), and albumin-to-alkaline phosphatase ratio (AAPR) and three metabolic dysfunction indicators, including TyG, TyG-BMI, and METS-IR were calculated based on preoperative laboratory examination. The detailed calculation formulas are provided in the Supplementary material. In addition, due to the limited availability of preoperative glucose and lipid biomarkers across the participating centers, metabolic-related indicators were only available in a subset of patients from center A. Therefore, a subcohort analysis was conducted to explore the associations between metabolic dysfunction indicators and OS.
SINRIs were analyzed both as quartiles, with the lowest quartile (Q1) used as the reference group, and as continuous variables per 1-standard deviation (SD) increase. Logistic regression analysis was performed to evaluate the associations between SINRIs and post-operative complications, and Cox proportional hazards (CoxPH) regression was used to evaluate the associations between SINRIs and OS. Model 1 was unadjusted. Model 2 was adjusted for age, sex, smoking status, alcohol consumption, hypertension, T2DM, and CHD. Model 3 was further adjusted for NAT, tumor location, and tumor stage. The results were presented as odds ratios or hazard ratios (HRs) with corresponding 95% confidence intervals (CIs). Restricted cubic spline (RCS) models were applied to assess dose-response relationships between SINRIs and the risk of post-operative complications and OS. In the center A sub-cohort with available metabolic data, TyG, TyG-BMI, and METS-IR were standardized and analyzed as continuous variables per 1-SD increase using the same regression framework and covariate-adjustment strategy.
To develop survival prediction models, the cohort from center A was used as the training dataset, whereas the cohorts from center B and the combined cohorts from centers C and D served as two independent external validation datasets. All 14 SINRIs and clinicopathological variables were subjected to feature selection using least absolute shrinkage and selection operator (LASSO) regression and the Boruta algorithm, and variables identified by both methods were retained for subsequent model development[22].
Six survival prediction models were constructed, including CoxPH, random survival forest (RSF), survival support vector machine (SSVM), gradient boosting survival analysis (GBSA), eXtreme gradient boosting for survival embeddings (XGBSE), and DeepSurv. Hyperparameters were optimized using grid search with 10-fold cross-validation in the training dataset.
Model performance was evaluated in the training and external validation cohorts. Discrimination was assessed using Harrell’s concordance index (C-index) and time-dependent receiver operating characteristic (ROC) analysis[23]. Calibration curves and decision curve analysis were further used to evaluate model calibration and clinical net benefit[24]. The model demonstrating the best overall performance across datasets was selected as the optimal model. Additional evaluation metrics, including probability density function (PDF), overlapping coefficient, integrated discrimination improvement (IDI), and continuous net reclassification improvement (cNRI) were calculated to compare the predictive performance between the optimal model and the conventional CoxPH[25].
To improve model interpretability, SHapley Additive exPlanations (SHAP) were applied to quantify the contribution of each predictor and to provide both global and local explanations of model predictions[26]. Risk stratification was performed based on the risk scores generated by the optimal model. The optimal cutoff value was determined using maximally selected rank statistics, and patients were categorized into high- and low-risk groups. Kaplan-Meier (KM) survival curves and log-rank tests were used to compare survival differences between groups[27]. Finally, the optimal model was deployed as an interactive application using the Streamlit framework to facilitate clinical implementation. The application generates individualized survival probabilities and cumulative risk curves based on patient-specific predictor values.
Continuous variables are presented as the median (interquartile range) for skewed distributions and compared using the Wilcoxon rank-sum test. Categorical variables were reported as n (%) and compared using the χ2 test. Continuous variables were standardized using z-score normalization to minimize the impact of feature scale differences. Variables with < 20% missing values were imputed using multiple imputation with predictive mean matching implemented in the mice package. All statistical analyses were two-sided, and a P < 0.05 was considered statistically significant. All analyses were performed using R (Version 4.4.1) and Python (Version 3.11). RSF, CoxPH, SSVM, and GBSA were developed using the scikit-survival package; XGBSE was implemented with the xgbse package; and DeepSurv was built using the pycox package.
A total of 10205 patients who underwent radical resection for CRC were included from four medical centers (6731 in center A, 2331 in center B, 481 in center C, and 656 in center D) (Figure 2). The median age of the study population was 65 years, including 5978 males and 4227 females. The median follow-up time was 48 months. The baseline clinical and laboratory characteristics of the study population are summarized in Table 1.
| Characteristic | n = 10205 |
| Sex | |
| Male | 5978 (58.6) |
| Female | 4227 (41.4) |
| Age | 65.00 (56.00, 72.00) |
| Smoking | 3630 (35.6) |
| Drinking | 3063 (30.0) |
| Hypertension | 2751 (27.0) |
| T2DM | 1305 (12.8) |
| CHD | 566 (5.5) |
| NAT | 665 (6.5) |
| Tumor location | |
| Rectal | 5273 (51.7) |
| Colon | 4932 (48.3) |
| Tumor stage | |
| I | 2092 (20.5) |
| II | 4089 (40.1) |
| III | 3504 (34.3) |
| IV | 520 (5.1) |
| Overall complications | 2212 (21.7) |
| Major complications | 179 (1.8) |
| Cardiovascular complications | 179 (1.8) |
| Digestive complications | 899 (8.8) |
| Respiratory complications | 643 (6.3) |
| Nutritional and metabolic complications | 571 (5.6) |
| Status | 1850 (18.1) |
| Follow-up (month) | 49.00 (28.00, 84.00) |
| CA19-9 | 11.39 (5.94, 22.60) |
| CEA | 3.53 (2.01, 8.10) |
| NLR | 2.61 (1.88, 3.79) |
| PLR | 157.89 (117.35, 216.85) |
| LMR | 3.48 (2.53, 4.71) |
| SII | 568.70 (373.61, 931.67) |
| SIRI | 1.03 (0.66, 1.69) |
| PIV | 221.67 (129.44, 409.69) |
| PNI | 47.35 (43.50, 51.00) |
| ALBI | -2.77 (-3.00, -2.50) |
| HALP | 31.47 (19.57, 45.73) |
| FAR | 0.09 (0.07, 0.10) |
| PAR | 5.45 (4.30, 6.93) |
| NPR | 0.02 (0.01, 0.02) |
| AGR | 1.44 (1.27, 1.65) |
| AAPR | 0.54 (0.45, 0.64) |
After multivariable adjustment, all 14 SINRIs were significantly associated with the risk of post-operative complications (Supplementary Table 1). In general, higher levels of NLR, PLR, SII, SIRI, PIV, ALBI, FAR, PAR, and NPR were associated with increased complication risk, whereas higher levels of LMR, PNI, HALP, AGR, and AAPR were associated with lower risk (Supplementary Table 1 and Figure 3). RCS analyses further suggested dose-response relationships between SINRIs and post-operative complications (Supplementary Figure 1). In addition, analyses according to complication severity and organ-system categories showed broadly consistent patterns, suggesting that the associations between SINRIs and post-operative complications were not limited to overall complications but were also observed across major and organ-specific complications (Figure 3 and Supplementary Table 2).
Higher levels of NLR, PLR, SII, SIRI, PIV, ALBI, FAR, and NPR were significantly associated with increased mortality risk, whereas higher levels of LMR, PNI, and HALP were associated with reduced mortality risk. By contrast, PAR, AGR, and AAPR were not significantly associated with OS after multivariable adjustment. Specifically, per 1-SD increase, NLR, PLR, SII, SIRI, PIV, ALBI, FAR, and NPR were associated with 15%, 12%, 13%, 10%, 10%, 16%, 15%, and 7% higher mortality risk, respectively. Conversely, LMR, PNI, and HALP were associated with 15%, 15%, and 10% lower mortality risk, respectively (Figure 4 and Supplementary Table 3). RCS analysis further supported dose-response relationships between SINRIs and mortality risk. Nonlinear associations were observed for NLR, LMR, SIRI, and PIV, whereas the remaining significant indicators showed predominantly linear associations with OS (Supplementary Figure 2).
A total of 1278 patients from center A were included in the subcohort analysis (Supplementary Table 4). Per 1-SD increase, TyG (HR = 1.263, 95%CI: 1.120-1.424), TyG-BMI (HR = 1.268, 95%CI: 1.108-1.451), and METS-IR (HR = 1.261, 95%CI: 1.108-1.436) were all significantly associated with poorer OS (Supplementary Figure 3A and Supplementary Table 5). RCS analysis further demonstrated linear associations between TyG, TyG-BMI, and METS-IR and mortality risk (all P for nonlinearity > 0.05), indicating that higher levels of these metabolic indicators were consistently associated with increased mortality risk (Supplementary Figure 3B).
A total of 6731 patients from center A were included in the training cohort, whereas 2331 patients from center B and 1137 patients from centers C and D were used as two independent external validation cohorts to evaluate the performance and stability of the models. Detailed data distributions across the three datasets are presented in Supplementary Table 6. All 14 SINRIs, together with demographic characteristics, tumor biomarkers, and clinicopathological variables, were included in the subsequent modeling. LASSO regression and the Boruta algorithm were then used for feature selection. Finally, tumor stage, age, NAT, carbohydrate antigen 19-9 (CA19-9), carcinoembryonic antigen (CEA), FAR, PNI, NLR, ALBI, and LMR were selected as the final input features for model development (Supplementary Figure 4). Hyper
The predictive performance of the six candidate models was evaluated in the training cohort and two external validation cohorts. In the training cohort, RSF, GBSA, SSVM, and XGBSE achieved higher C-index values than the CoxPH model, with RSF and GBSA showing the highest time-dependent area under the curves (AUCs) for 1-year, 3-tear, and 5-year survival prediction (Figure 5A and B).
In the external validation cohorts, model performance decreased but remained relatively stable (Figure 6). Most models achieved C-index values around 0.75 in external validation cohort 1 (Figure 6A and B). In external validation cohort 2, GBSA, RSF, and SSVM maintained comparable discrimination and stable time-dependent AUCs for 1-year, 3-year, and 5-year survival prediction (Figure 6E and F).
Although several ML models showed comparable discrimination in external validation, GBSA demonstrated the most favorable overall performance in terms of calibration, clinical net benefit, and risk reclassification. Specifically, GBSA showed calibration curves closest to the 45° reference line, provided greater clinical net benefit across a wide range of threshold probabilities, and achieved improved discrimination and risk reclassification compared with CoxPH, as indicated by PDF plots, IDI, and cNRI analyses (Figure 5C and D, Figure 6C, D, G and H, Supplementary Figures 5 and
SHAP summary plots ranked predictors according to their mean absolute SHAP values, providing a global interpretation of the final GBSA model. Tumor stage, age, CA19-9, CEA, FAR, and overall complications emerged as the six most influential variables, with tumor stage ranked first. SHAP dependence plots further illustrated risk thresholds for the top predictors. Patients with tumor stage > II, age > 66 years, CA19-9 > 27.97 ng/mL, CEA > 6.01 ng/mL, NLR > 2.73, and FAR > 0.09 were associated with a higher predicted mortality risk, suggesting the need for enhanced clinical surveillance or intervention (Figure 7).
The optimal cutoff value for the risk score was 0.77 (Supplementary Figure 7A). Patients were subsequently classified into high and low risk groups according to this threshold. KM survival analysis showed clear separation between the two groups, confirming effective prognostic stratification across training and external validation cohorts (Supplementary Figure 7B-D).
To facilitate clinical implementation, the final GBSA model was deployed as an interactive web-based prediction tool using the Streamlit Python framework (available from: https://survival-prediction-crc.streamlit.app/). After inputting the clinical characteristics of an individual patient undergoing radical colectomy, the application provides predicted 1-year, 3-year, and 5-year survival probabilities and automatically classifies the patient into risk groups. Additionally, an individualized survival curve is generated to visualize the time-dependent mortality risk.
In this multicenter cohort study of 10205 patients undergoing radical surgery for CRC across four high-volume hospitals in Southwest China, we systematically evaluated the prognostic value of 14 inflammation- and nutrition-related indicators in CRC. In addition, we performed a sub-cohort analysis to explore the associations of metabolic dysfunction indicators, including TyG, TyG-BMI, and METS-IR, with OS. Furthermore, we developed and externally validated ML-based models for survival prediction and translated the final optimal model into a clinically applicable decision-support tool.
Among the 14 SINRIs, all were significantly associated with post-operative complication risk. Higher levels of inflammatory indicators and lower levels of protective markers such as LMR, PNI, and HALP were associated with worse clinical outcomes. For long-term survival, several composite indicators, including LMR, SII, PNI, and HALP, remained independent predictors of mortality, highlighting the important role of host inflammatory and nutritional status in CRC progression and long-term prognosis. These findings are broadly consistent with previous studies showing that several blood-based inflammatory or nutritional markers are associated with adverse outcomes in CRC[9,10,28]. In particular, NLR, LMR, SIRI, ALBI, and PNI have been extensively investigated as prognostic indicators in gastrointestinal malignancies. For example, NLR and SIRI mainly reflect the balance between systemic inflammatory activation and anti-tumor immune surveillance. Elevated neutrophils and monocytes may promote tumor progression through secretion of pro-inflammatory cytokines, support of angiogenesis, and suppression of anti-tumor immunity, whereas reduced lymphocyte counts may indicate weakened immune defense[29,30]. In contrast, ALBI and PNI reflect nutritional condition, hepatic synthetic function, and overall physiological reserve. Poor nutritional or hepatic status may impair tolerance to surgery and systemic treatment, thereby contributing to poorer recovery and survival[31,32].
In the exploratory sub-cohort analysis, TyG, TyG-BMI, and METS-IR were significantly associated with worse OS, and spline analyses suggested approximately linear dose-response relationships. These findings are consistent with the hypothesis that insulin resistance-related metabolic disturbances may be associated with CRC progression and adverse outcomes. Dysregulated glucose-lipid metabolism may contribute to chronic low-grade inflammation, oxidative stress, immune dysregulation, and impaired nutritional-metabolic homeostasis, all of which are closely associated with CRC progression[33-35]. However, the metabolic dysfunction indicators were assessed in a single-center sub-cohort with a limited sample size and were not externally validated. Therefore, these findings should be considered exploratory and interpreted with caution. Future multi-center prospective cohorts are needed to validate their prognostic value in CRC.
Furthermore, we integrated SINRIs with clinicopathological variables characteristics to construct and externally validate six survival prediction models. Among them, the GBSA model demonstrated the most stable performance across the multi-center datasets. Compared with the conventional CoxPH model, GBSA achieved improved discrimination and risk reclassification, while also maintaining favorable calibration and clinical net benefit. The SHAP analysis highlighted tumor stage, age, CA19-9, CEA, NLR, and FAR as key predictors of survival, which is largely consistent with the well-established CRC prognostic model[36-38]. Notably, FAR contributed meaningfully to survival prediction. As FAR is derived from fibrinogen and albumin, an elevated FAR may reflect an active inflammatory and pro-coagulant state together with impaired nutritional reserve. This biological profile may facilitate tumor progression, support circulating tumor cell survival, and disrupt vascular stability, thereby contributing to poorer outcomes[39,40].
This study has several notable strengths. First, the evaluated biomarkers are inexpensive and routinely available in clinical practice, facilitating their potential application in peri-operative risk assessment. Second, by simultaneously integrating inflammatory, nutritional, and metabolic information, our approach provides a more comprehensive evaluation of the systemic condition of patients. Third, our study included a large multi-center cohort, systematically compared a broad panel of composite indicators, and combined conventional regression analysis with ML modeling and external validation, enhancing the robustness and generalizability of the findings. Fourth, the development of a web-based prediction tool further improves the clinical applicability of the model and may facilitate individualized follow-up and post-operative management in routine practice.
Nevertheless, several limitations should be acknowledged. First, this was a retrospective cohort study, and residual confounding may remain despite multivariable adjustment. Second, although the study was multi-center, all participating centers were located in Southwest China, which may limit the geographic and ethnic generalizability of the findings to other populations. Third, the exploratory analysis of metabolic dysfunction indicators was limited by its single-center design and lack of external validation. Fourth, the present study lacked dynamic biomarker measurements, as only preoperative measurements were analyzed. Fifth, although the GBSA model showed stable performance, its external discriminative performance remained moderate. Therefore, the model may be more appropriately used as an auxiliary risk stratification tool together with established clinicopathological factors and clinical judgment. Further refine
In conclusion, this multicenter study demonstrated that SINRIs were significantly associated with post-operative complications and OS in patients with CRC undergoing radical surgery. Exploratory sub-cohort analyses suggested that metabolic dysfunction indicators may provide complementary prognostic information. The GBSA-based ML model showed stable multicenter performance and effective risk stratification, supporting its potential utility as an auxiliary tool for individualized prognostic assessment. These findings highlight the value of integrating multidimensional blood-based biomarkers with ML methods to improve risk stratification in CRC.
We acknowledge all authors whose publications are referred to in our article.
| 1. | Li Q, Geng S, Luo H, Wang W, Mo YQ, Luo Q, Wang L, Song GB, Sheng JP, Xu B. Signaling pathways involved in colorectal cancer: pathogenesis and targeted therapy. Signal Transduct Target Ther. 2024;9:266. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 398] [Cited by in RCA: 428] [Article Influence: 214.0] [Reference Citation Analysis (16)] |
| 2. | Cienfuegos JA, Baixauli J, Beorlegui C, Ortega PM, Granero L, Zozaya G, Hernández Lizoáin JL. The impact of major postoperative complications on long-term outcomes following curative resection of colon cancer. Int J Surg. 2018;52:303-308. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 13] [Cited by in RCA: 23] [Article Influence: 2.9] [Reference Citation Analysis (1)] |
| 3. | Arnold M, Sierra MS, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global patterns and trends in colorectal cancer incidence and mortality. Gut. 2017;66:683-691. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3978] [Cited by in RCA: 3657] [Article Influence: 406.3] [Reference Citation Analysis (23)] |
| 4. | Schmitt M, Greten FR. The inflammatory pathogenesis of colorectal cancer. Nat Rev Immunol. 2021;21:653-667. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 644] [Cited by in RCA: 591] [Article Influence: 118.2] [Reference Citation Analysis (10)] |
| 5. | Nishida A, Andoh A. The Role of Inflammation in Cancer: Mechanisms of Tumor Initiation, Progression, and Metastasis. Cells. 2025;14:488. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 157] [Cited by in RCA: 191] [Article Influence: 191.0] [Reference Citation Analysis (2)] |
| 6. | Arends J, Baracos V, Bertz H, Bozzetti F, Calder PC, Deutz NEP, Erickson N, Laviano A, Lisanti MP, Lobo DN, McMillan DC, Muscaritoli M, Ockenga J, Pirlich M, Strasser F, de van der Schueren M, Van Gossum A, Vaupel P, Weimann A. ESPEN expert group recommendations for action against cancer-related malnutrition. Clin Nutr. 2017;36:1187-1196. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1037] [Cited by in RCA: 901] [Article Influence: 100.1] [Reference Citation Analysis (10)] |
| 7. | Bossi P, Delrio P, Mascheroni A, Zanetti M. The Spectrum of Malnutrition/Cachexia/Sarcopenia in Oncology According to Different Cancer Types and Settings: A Narrative Review. Nutrients. 2021;13:1980. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 177] [Cited by in RCA: 362] [Article Influence: 72.4] [Reference Citation Analysis (4)] |
| 8. | Li J, Zhu N, Wang C, You L, Guo W, Yuan Z, Qi S, Zhao H, Yu J, Huang Y. Preoperative albumin-to-globulin ratio and prognostic nutritional index predict the prognosis of colorectal cancer: a retrospective study. Sci Rep. 2023;13:17272. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 69] [Cited by in RCA: 85] [Article Influence: 28.3] [Reference Citation Analysis (3)] |
| 9. | Menyhart O, Fekete JT, Győrffy B. Inflammation and Colorectal Cancer: A Meta-Analysis of the Prognostic Significance of the Systemic Immune-Inflammation Index (SII) and the Systemic Inflammation Response Index (SIRI). Int J Mol Sci. 2024;25:8441. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 86] [Reference Citation Analysis (4)] |
| 10. | Kim D, Lee JH, Cho ES, Shin SJ, Lee HS, Koh HH, Lee KY, Kang J. Clinical Significance of Combining Preoperative and Postoperative Albumin-Bilirubin Score in Colorectal Cancer. Cancer Res Treat. 2023;55:1261-1269. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 12] [Article Influence: 4.0] [Reference Citation Analysis (0)] |
| 11. | Peng D, Li ZW, Liu F, Liu XR, Wang CY. Predictive value of red blood cell distribution width and hematocrit for short-term outcomes and prognosis in colorectal cancer patients undergoing radical surgery. World J Gastroenterol. 2024;30:1714-1726. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 11] [Reference Citation Analysis (2)] |
| 12. | Faubert B, Solmonson A, DeBerardinis RJ. Metabolic reprogramming and cancer progression. Science. 2020;368:eaaw5473. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 2570] [Cited by in RCA: 2096] [Article Influence: 349.3] [Reference Citation Analysis (7)] |
| 13. | Ma Z, Wang S, Liu S, Yang W, Hu J, Lv L, Yu Q, Lu Y. Metabolic syndrome in colorectal cancer liver metastasis: metabolic reprogramming and microenvironment crosstalk. Front Immunol. 2025;16:1653442. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 10] [Reference Citation Analysis (0)] |
| 14. | Cai S, Xing H, Wang Y, Yang W, Luo H, Ye X. The association between triglyceride glucose-body mass index and overall survival in postoperative patient with lung cancer. Front Endocrinol (Lausanne). 2025;16:1528644. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 3] [Reference Citation Analysis (2)] |
| 15. | Meng HJ, Mao Y, Zhao DQ, Shi SG. Prognostic value of the triglyceride-glucose index in advanced gastric cancer: A call for further exploration. World J Gastroenterol. 2025;31:104574. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 1] [Reference Citation Analysis (6)] |
| 16. | Song Y, Jiang L, Han Y, Zhang S, Li S. Triglyceride-glucose index and glycemic dynamics in pancreatic ductal adenocarcinoma: implications for disease progression and prognosis. J Transl Med. 2024;22:708. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 14] [Reference Citation Analysis (2)] |
| 17. | Jiang N, Wu Y, Li C. Limitations of using COX proportional hazards model in cardiovascular research. Cardiovasc Diabetol. 2024;23:219. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 16] [Reference Citation Analysis (0)] |
| 18. | Kim HU, Han JW, Sung PS, Jang JW, Yoon SK, Choi HJ, You YK. Machine Learning-Based Selection of Resection vs Transplant and Survival in Hepatocellular Carcinoma. JAMA Netw Open. 2025;8:e2532353. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 4] [Cited by in RCA: 4] [Article Influence: 4.0] [Reference Citation Analysis (0)] |
| 19. | Hessulf F, Bhatt DL, Engdahl J, Lundgren P, Omerovic E, Rawshani A, Helleryd E, Dworeck C, Friberg H, Redfors B, Nielsen N, Myredal A, Frigyesi A, Herlitz J, Rawshani A. Predicting survival and neurological outcome in out-of-hospital cardiac arrest using machine learning: the SCARS model. EBioMedicine. 2023;89:104464. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 4] [Cited by in RCA: 46] [Article Influence: 15.3] [Reference Citation Analysis (0)] |
| 20. | He X, Wu W, Lin Z, Ding Y, Si J, Sun LM. Validation of the American Joint Committee on Cancer (AJCC) 8th edition stage system for gastric cancer patients: a population-based analysis. Gastric Cancer. 2018;21:391-400. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 53] [Cited by in RCA: 57] [Article Influence: 7.1] [Reference Citation Analysis (19)] |
| 21. | Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205-213. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 18949] [Reference Citation Analysis (0)] |
| 22. | Alba AC, Agoritsas T, Walsh M, Hanna S, Iorio A, Devereaux PJ, McGinn T, Guyatt G. Discrimination and Calibration of Clinical Prediction Models: Users' Guides to the Medical Literature. JAMA. 2017;318:1377-1384. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1323] [Cited by in RCA: 1270] [Article Influence: 141.1] [Reference Citation Analysis (7)] |
| 23. | Yuan L, Wang L, Gao J, Chen X, Wang H, Tan WS, Sun K, Gong Y, Deng W. OncoE25: an AI model for predicting postoperative prognosis in early-onset stage I-III colon and rectal cancer-a population-based study using SEER with dual-center cohort validation. J Transl Med. 2025;23:695. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 5] [Cited by in RCA: 6] [Article Influence: 6.0] [Reference Citation Analysis (0)] |
| 24. | Luo J, Liang X, Xin J, Li J, Li P, Zhou Q, Hao S, Zhang H, Lu Y, Wu T, Yang L, Li J, Li T, Ren K, Guo B, Zhou X, Chen J, He L, Yang H, Hu W, You S, Xin S, Jiang J, Shi D, Chen X, Li J; Chinese Group on the Study of Severe Hepatitis B (COSSH). Predicting the Onset of Hepatitis B Virus-Related Acute-on-Chronic Liver Failure. Clin Gastroenterol Hepatol. 2023;21:681-693. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 3] [Cited by in RCA: 31] [Article Influence: 10.3] [Reference Citation Analysis (0)] |
| 25. | Dong R, Luo Z, Xue H, Shao J, Chen L, Jin W, Yang L, Shen C, Xu M, Wu M, Wang J. Development and Validation of an Explainable Machine Learning Model for Warning of Hepatitis E Virus-Related Acute Liver Failure. Liver Int. 2025;45:e70129. [RCA] [PubMed] [DOI] [Full Text] [Cited by in Crossref: 1] [Cited by in RCA: 2] [Article Influence: 2.0] [Reference Citation Analysis (0)] |
| 26. | Li GH, Cheung CL, Tan KC, Kung AW, Kwok TC, Lau WC, Wong JS, Hsu WWQ, Fang C, Wong IC. Development and validation of sex-specific hip fracture prediction models using electronic health records: a retrospective, population-based cohort study. EClinicalMedicine. 2023;58:101876. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 24] [Reference Citation Analysis (0)] |
| 27. | Rui F, Yeo YH, Xu L, Zheng Q, Xu X, Ni W, Tan Y, Zeng QL, He Z, Tian X, Xue Q, Qiu Y, Zhu C, Ding W, Wang J, Huang R, Xu Y, Chen Y, Fan J, Fan Z, Qi X, Huang DQ, Xie Q, Shi J, Wu C, Li J. Development of a machine learning-based model to predict hepatic inflammation in chronic hepatitis B patients with concurrent hepatic steatosis: a cohort study. EClinicalMedicine. 2024;68:102419. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 48] [Cited by in RCA: 40] [Article Influence: 20.0] [Reference Citation Analysis (4)] |
| 28. | Zhou M, Gu Q, Zhou M, Yang S, Liu Y, Guan B, Xie B, Han A, Xiang J, Yan D. Extensive study on the associations of 12 composite inflammatory indices with colorectal cancer risk and mortality: a cross-sectional analysis of NHANES 2001-2020. Int J Surg. 2025;111:7559-7575. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 7] [Cited by in RCA: 26] [Article Influence: 26.0] [Reference Citation Analysis (0)] |
| 29. | Yang FR, Li HL, Hu XW, Sha R, Li HJ. Association between the systemic inflammation response index and mortality in cancer survivors based on NHANES 2001-2018. Sci Rep. 2025;15:15151. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 6] [Reference Citation Analysis (1)] |
| 30. | Ye K, Xiao M, Li Z, He K, Wang J, Zhu L, Xiong W, Zhong Z, Tang Y. Preoperative systemic inflammation response index is an independent prognostic marker for BCG immunotherapy in patients with non-muscle-invasive bladder cancer. Cancer Med. 2023;12:4206-4217. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 20] [Cited by in RCA: 18] [Article Influence: 6.0] [Reference Citation Analysis (1)] |
| 31. | Su Z, Lin Y, Li M, Yang Y, Chen X, Zhu Y, Mo Y, Huang Z, Liu Y, Michael N. Preoperative nutritional status as a predictor of postoperative overall survival in abdominal tumor surgery: a systematic review and meta-analysis. Front Surg. 2025;12:1645392. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1] [Cited by in RCA: 3] [Article Influence: 3.0] [Reference Citation Analysis (0)] |
| 32. | Zhang Y, Tang M, Gu QH, Zhou LN, Chen MB. The Prognostic Value of Combined Systemic Immune-Inflammatory Index (SII) and Prognostic Nutritional Index (PNI) in Solid Tumor. Cancer Manag Res. 2025;17:1351-1359. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in RCA: 9] [Reference Citation Analysis (4)] |
| 33. | Omer HFE, Alghazali M, Ibrahim MY, Abdalla NMY, Hassan AHM, Yousif EAS, Abdhameed AEB, Naser YWS, Hamad NME, Abdalla MAI, Idres MOM, Ahmed AAO, Elhadi YAM, Mohamed SOO. Association between the triglyceride-glucose index (TyG Index) and risk of colorectal cancer: a systematic review and meta-analysis. World J Surg Oncol. 2025;23:280. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 6] [Cited by in RCA: 14] [Article Influence: 14.0] [Reference Citation Analysis (1)] |
| 34. | Luo L, Liu Z, Gan Q, Feng J, Wang L, Ouyang W. Relationships of triglyceride-glucose-related indices with colorectal cancer incidence and mortality in an American population: a dose-response meta-analysis and cohort study. Lipids Health Dis. 2025;24:151. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 1] [Cited by in RCA: 9] [Article Influence: 9.0] [Reference Citation Analysis (0)] |
| 35. | Liu XY, Shu XP, Shen Y, Hai ZX, Wang L, Xiang CY, Zhao JN, Li SQ, Kang B, Peng D. Peri-operative metabolomic alterations predict post-operative complications and hospital stay in colorectal cancer patients. Int J Surg. 2025. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 1] [Reference Citation Analysis (0)] |
| 36. | Yang X, Qiu H, Wang L, Wang X. Predicting Colorectal Cancer Survival Using Time-to-Event Machine Learning: Retrospective Cohort Study. J Med Internet Res. 2023;25:e44417. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 27] [Cited by in RCA: 44] [Article Influence: 14.7] [Reference Citation Analysis (4)] |
| 37. | Cui Y, Zhao K, Meng X, Mao Y, Han C, Shi Z, Yang X, Tong T, Wu L, Liu Z. A computed tomography-based multitask deep learning model for predicting tumour stroma ratio and treatment outcomes in patients with colorectal cancer: a multicentre cohort study. Int J Surg. 2024;110:2845-2854. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 4] [Cited by in RCA: 15] [Article Influence: 7.5] [Reference Citation Analysis (0)] |
| 38. | Buk Cardoso L, Cunha Parro V, Verzinhasse Peres S, Curado MP, Fernandes GA, Wünsch Filho V, Natasha Toporcov T. Machine learning for predicting survival of colorectal cancer patients. Sci Rep. 2023;13:8874. [RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)] [Cited by in Crossref: 3] [Cited by in RCA: 28] [Article Influence: 9.3] [Reference Citation Analysis (0)] |
| 39. | Wang Y, Schneider SW, Gorzelanny C. Crosstalk between Circulating Tumor Cells and Plasma Proteins-Impact on Coagulation and Anticoagulation. Cancers (Basel). 2023;15:3025. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 6] [Reference Citation Analysis (0)] |
| 40. | Wu X, Yu X, Chen C, Chen C, Wang Y, Su D, Zhu L. Fibrinogen and tumors. Front Oncol. 2024;14:1393599. [RCA] [PubMed] [DOI] [Full Text] [Cited by in RCA: 22] [Reference Citation Analysis (2)] |