Development and validation of a nomogram comprising GLUT1 and clinical characteristics for predicting overall survival in intrahepatic cholangiocarcinoma

Introduction

Cholangiocarcinoma (CCA) is the second most common type of hepatic malignancy, which has increasing incidence and caused serious public health problems over the past decade (1,2). According to anatomic location, CCA can be divided into intrahepatic (iCCA), perihilar (pCCA) and distal (dCCA) (3). Importantly, there are substantial differences among the anatomical CCA subtypes in terms of incidence rate, biological characteristics and prognosis. In particular, the incidence rate of iCCA is increasing rapidly in China and surgical resection is not usually possible due to most patients are at an advanced stage at the time of diagnosis (4,5). Clarifying the molecular mechanisms of iCCA is key to identify novel therapeutic targets and predictive biomarkers.

Regarding malignant tumors, one of the hallmarks is that glucose metabolism in cells is reprogramed to support the energy supply needed for cancer cell growth and proliferation (6). The increased glucose uptake depends on glucose transporters (GLUTs) expressed on the cell membrane (7). Among the GLUT protein family, GLUT1 plays a key role in glucose uptake and is abnormally overexpressed in various cancers such as hepatocellular carcinoma (8), colorectal cancer (9), and pancreatic cancer (10). Recently, it has been reported that GLUT1 is upregulated in iCCA tissues and could enhance the migration and invasion ability of iCCA cells (11,12). However, to date, only a few studies have explored the role of GLUT1 in iCCA, and the clinical significance and prognostic value of GLUT1 in iCCA remain unclear.

Nomogram is a decision-making tool, which is widely used in predicting prognosis of various cancers (13). Nomogram can estimate patient outcomes effectively and accurately by incorporating multiple clinicopathological variables and several important prognostic factors (14). Moreover, nomograms are graphical calculators that use graduated scales to predict outcome probabilities (15). Therefore, with user-friendly digital interface, higher accuracy and more intuitive prognosis model, nomogram can be more helpful in clinical prognosis decision-making.

Current iCCA prognostic models primarily rely on clinicopathological factors like tumor-node-metastasis (TNM) staging or imaging-based tools such as upper abdominal enhanced computed tomography (CT) scan predicts lymph node metastasis (16). However, these exhibit suboptimal accuracy, area under the curve (AUC) <0.70 and fail to incorporate molecular drivers of progression. While iCCA is an aggressive malignancy, and its progression is influenced by a variety of clinical features such as age, tumor size and biological factors (17,18). Nomogram can integrate multiple clinicopathological factors and emerging biomarkers to stratify prognosis, offering a comprehensive model for assessing iCCA patient outcomes. To date, no model integrates metabolic markers like GLUT1 with clinical variables for iCCA outcome prediction.

Our study addresses these gaps by exploring GLUT1’s prognostic value in iCCA and developing a comprehensive nomogram. By combining GLUT1 expression with clinical factors such as age, tumor size, we aim to create a robust tool for predicting 2- and 3-year overall survival (OS), improving risk stratification and guiding treatment strategies. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-630/rc).

Methods

Ethical approval

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institute Research Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University (No. SL-II2024-355-01) and informed consent was obtained from all individual participants.

Patient samples

Sample size was determined a priori using the R pmsampsize package, indicating ≥85 patients provided 80% power [α=0.05, two-sided, hazard ratio (HR) ≥1.8] for Cox regression with 3 predictors. From April 1, 2017 to April 1, 2023, paraffin-embedded paired tumor and non-tumor tissues were collected from 95 patients who had iCCA resection at The Third Affiliated Hospital of Sun Yat-sen University. Clinical variables such as age, tumor size, TNM stage, vascular invasion were extracted from the hospital record.

Postoperative follow-up was conducted per institutional protocols: quarterly abdominal contrast-enhanced CT/magnetic resonance imaging (MRI) and CA19-9 measurements for the first 2 years, followed by biannual assessments. The median follow-up duration was 32 months (range, 6–78 months), with 52 (54.7%) patients experiencing recurrence and 41 (43.2%) deaths. OS was calculated as the time from the surgery date until death or the last follow-up. All patients were staged following the eighth American Joint Committee on Cancer (AJCC) TNM staging system for CCA (19).

TIMER and UALCAN database analysis

The differential expression levels of SLC2A1 between tumor and adjacent normal tissue across different types of cancer from The Cancer Genome Atlas (TCGA) database were analyzed by TIMER (http://timer.cistrome.org/). And we used the UALCAN database (http://ualcan.path.uab.edu) to examine the variation in SLC2A1 mRNA expression between tumor and adjacent normal tissue in iCCA.

Immunohistochemistry (IHC) staining and quantification

Sections of 4 µm thickness were cut from the paraffin-embedded tissues. Then, the IHC analysis was performed as per the previously detailed method (20), with Dako REAL^TM detection systems (Dako, Glostrup, Denmark, Cat No. K5007).

Briefly, the tissue sections were dewaxed in xylene, rehydrated in gradient ethanol solutions, and then endogenous peroxidase activity was blocked with hydrogen peroxide. Subsequently heat-mediated antigen retrieval in citrate buffer (pH 6.0) was used for 10 minutes. Sections were incubated with antibody specific for GLUT1 (1:500 dilution, Abcam, Cambridge, UK, Cat No. ab115730) and at 4 ℃ overnight, followed by incubation with secondary antibody for 30 minutes. IHC was visualized with diaminobenzidine (DAB) and brown color suggested positive staining.

To evaluate the expression level of GLUT1, an automatic slide scanner (Pannoramic MIDI, 3DHISTECH Ltd., Budapest, Hungary) were performed to obtain a whole digital image of the immunohistochemical slides [high-power fields (HPFs), 20×10] and 5 microscopic fields from tumor or peritumor regions were randomly selected using 3DHISTECH Panoramic Viewer. Then the Inform image analysis software (version 2.0.1; Perkin-Elmer Applied Biosystems, Foster City, USA) was applied for image quantification as described in our previous study (21). Finally, the object density (per megapixel) was counted by quantifying the image signals of the selected region of interest according to the manufacturer’s protocol.

Nomogram construction and validation

We built a nomogram relying on the outcomes of independent risk factors in multivariable analyses for the prediction of 2- and 3-year OS rates. A Cox proportional hazards regression model was built on these independent risk factors using the R package rms. Due to the limited availability of external cohorts during the study period, internal validation was performed via 500-time bootstrap resampling to mitigate overfitting. To comprehensively assess the performance of the nomogram model, the survival receiver operating characteristic (ROC) package in R was employed to conduct a time-dependent ROC analysis and AUC >0.70 was predefined as indicating clinically useful discrimination. Furthermore, decision curve analysis (DCA) was utilized. This technique is crucial for determining whether clinical decisions informed by the nomogram could lead to improved patient outcomes, thereby evaluating the practical utility of the nomogram in a real-world clinical setting (22). DCA is a powerful tool for assessing the clinical utility of prediction models by considering the threshold probability, which represents the point at which the harm of a false-positive intervention exceeds that of a false-negative non-intervention. To evaluate the clinical value of the nomogram, we conducted a DCA by calculating the net benefits across a range of threshold probabilities. All statistical analyses were performed using R software (version 4.0.0; http://www.r-project.org/).

Statistical analysis

Statistical analyses were performed with SPSS 21.0 (IBM Corporation, Armonk, NY, USA) or GraphPad Prism5. The significance of differences between two groups were analyzed using the Mann-Whitney U test or Wilcoxon matched-pairs rank test and the correlation between GLUT1 expression and clinicopathological features was analyzed by Chi-squared test or Fisher’s exact test, as indicated. Prognostic factors associated with survival by univariate analysis were adopted as covariates in multivariate Cox proportional hazards analysis. Survival analysis was carried out using the Kaplan-Meier approach, and the log-rank test was performed to check for significance. All P values were two-sided, statistical significance was set at P<0.05.

Results

Profiling of SLC2A1 in various tumors

Increasing evidences indicate that GLUT1 expression is upregulated in a variety of cancers and correlates with poor prognosis, which implicates a new target or biomarker for cancer diagnosis (23). Here, we first evaluated the mRNA expression levels of SLC2A1 in different tumors and adjacent normal tissues by TCGA database. As shown in Figure 1, compared with adjacent normal tissues, SLC2A1 expression levels were significantly elevated in various cancerous tissues such as breast invasive carcinoma (BRCA), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), kidney chromophobe (KICH), kidney renal clear cell carcinoma (KIRC), liver hepatocellular carcinoma (LIHC) and so on. The expression profiling of the mRNA levels of SLC2A1 suggested that it may be acting as an important oncogenic molecule for tumor progression.

Figure 1 The expression of SLC2A1 in the tumor tissues and in adjacent normal tissues. The SLC2A1 expression was analyzed in various tumor tissues and adjacent normal tissues through the TCGA database (**, P<0.01; ***, P<0.001). TCGA, The Cancer Genome Atlas; TPM, transcripts per million.

GLUT1 expression and its correlation with clinicopathological characteristics and prognosis in iCCA

We further analyzed the protein expression of GLUT1 in 95 paraffin-embedded iCCA samples by IHC. There was distinguishable membrane staining of GLUT1 in iCCA tumor tissue, whereas the staining was weak in the adjacent normal tissue (Figure 2A). As shown in Figure 2B, the expression of GLUT1 quantified by the object density (per megapixel) was significantly increased in tumor tissues (median score =962.08) than in adjacent non-tumor tissues (median score =109.90; P<0.001). Furthermore, analysis of the mRNA levels of SLC2A1 expression in 32 CCA patients (iCCA: n=28; pCCA: n=3; dCCA: n=2) and 9 healthy subjects by UALCAN database gave similar result (median score: tumor tissues =10.213, normal tissue =0.878; P<0.05, Figure 2C).

Figure 2 Significantly elevated GLUT1 expression levels in iCCA and its correlation with clinicopathological characteristics and prognosis. (A) IHC assays of GLUT1 expression in tumor [T] and adjacent non-tumor tissues [N]. Scale bars =50 µm. The red arc represents the junction between tumor and adjacent non-tumor tissues, and the image on the right is an enlargement of the red box in the left image. (B) The object densities for GLUT1 expression of tumor [T] tissues are significantly higher than adjacent non-tumor tissues [N]. Data are shown as boxplot. (C) The mRNA levels of SLC2A1 expression in 32 CCA patients and 9 healthy subjects in the UALCAN database. (D-G) The object densities of GLUT1 expression are significantly correlated with vascular invasion (D), advanced T stage (E), advanced N stage (F) and advanced TNM stage (G). (H) iCCA patients with higher GLUT1 expression have a worse OS. *, P<0.05; **, P<0.01; ***, P<0.001. CCA, cholangiocarcinoma; iCCA, intrahepatic cholangiocarcinoma; IHC, immunohistochemistry; N, node; OS, overall survival; T, tumor; TCGA, The Cancer Genome Atlas; TNM, tumor-node-metastasis.

The clinicopathological characteristics of the cohort of 95 iCCA patients were summarized in Table 1. Subsequently, patients were classified in accordance with the object density of GLUT1. ROC analysis was utilized to discriminate between high and low GLUT1 expression, with the cut-off value for GLUT1 expression fixed at 750.18 (P=0.02). As shown in Table 2, high GLUT1 expression was significantly correlated with vascular invasion (P=0.004), advanced T stage (P=0.046), advanced N stage (P=0.01) and advanced TNM stage (P=0.004). And patients with vascular invasion (P=0.043, Figure 2D), advanced T stage (P=0.02, Figure 2E), advanced N stage (P=0.005, Figure 2F) and advanced TNM stage (P=0.006, Figure 2G) tended to have higher expression of GLUT1. Additionally, iCCA patients with high GLUT1 expression had a worse OS (P=0.001, Figure 2H). Results above suggested that GLUT1 is a promising prognostic predictor for iCCA patients.

Prognostic factors of iCCA

To comprehensively investigate the prognostic factors of iCCA, GLUT1 and clinicopathological factors were incorporated to filter predictors by univariate analysis. As presented in Table 3, age, tumor size, vascular invasion, T stage, N stage, TNM stage, and GLUT1 were associated with OS in iCCA. And these variables were further selected for the multivariate model, according to the Cox proportional hazards model (Table 3), three factors were independently correlated with prognosis, including age (P=0.02; HR, 1.002; 95% CI: 1.000 to 1.003), tumor size (P=0.03; HR, 1.118; 95% CI: 1.010 to 1.238) and GLUT1 (P=0.04; HR, 1.001; 95% CI: 1.000 to 1.002).

Development of an individualized predictive nomogram for iCCA

The above three independent prognostic factors were further utilized to establish the nomogram, based on a Cox regression model. We estimated the 2- and 3-year survival probabilities for iCCA by summing up the scores of each selected variable (Figure 3A). Patients with a higher score (total points >100) had a worse prognosis, compared with patients with a lower score (total points ≤100; P<0.001, Figure 3B). Moreover, the AUC of the nomogram for 2- and 3-year prediction was significantly higher than that of the TNM staging system (2-year, 0.74; 95% CI: 0.63 to 0.85 vs. 0.69; 95% CI: 0.58 to 0.80, respectively; 3-year, 0.77; 95% CI: 0.65 to 0.89 vs. 0.67; 95% CI: 0.55 to 0.79, respectively; Figure 3C), which indicated that the nomogram had better discriminatory power compared to the conventional TNM staging system for OS of iCCA.

Figure 3 Nomogram for predicting the 2- and 3-year overall survival rates of patients with iCCA. (A) Nomogram for predicting OS of patients with iCCA after resection. (B) Patient with a higher total point (>100 points) had a worse OS in iCCA. (C) Time-dependent ROC curve analysis of the OS nomogram predicting 2- and 3-year survival in the cohort compared with TNM staging. The AUC is indicated as the mean and 95% CI, estimated using 500× bootstrap resampling. AUC, area under the curve; CI, confidence interval; iCCA, intrahepatic cholangiocarcinoma; OS, overall survival; ROC, receiver operating characteristic; TNM, tumor-node-metastasis.

Validation and clinical use of the predictive nomogram

Calibration analysis was performed to confirm the resemblance between the survival rates predicted by the nomogram and the real-world survival rates. As shown by the calibration curve, there was excellent agreement between what the model predicted for 2- and 3-year OS and the real-life outcomes (Figure 4A,4B). In addition, the DCA was performed to analyze the clinical usefulness of the predictive nomogram (24,25). According to the DCA, the nomogram demonstrated the net advantage of decisions made with the help of the nomogram over a wide spectrum of threshold probabilities (Figure 4C), and showed superior discriminatory ability compared to the traditional TNM staging system in terms of 3-year OS for medical decisions after iCCA resection (Figure 4D).

Figure 4 Validation and clinical use of the predictive nomogram. (A,B) Calibration plots of 2-year (A) and 3-year (B) survival rates for the nomogram. The x-axis represents the nomogram predicted survival rate, whereas the y-axis represents the actual survival rate. The 45-degree line shows perfect calibration when predicted probabilities coincide with the actual outcomes. (C) DCA estimates the clinical usefulness for the OS nomogram. The solid line represents the nomogram predicted survival rate. The colored dotted line represents the assumption that all patients are subject to intervention. The dotted line in grey represents the assumption that no patients are subject to intervention. The colored solid lines represent the assumption that intervention was conducted based on the nomogram prediction. The y-axis measures the net benefit. (D) DCA of the OS nomogram predicting 3-year survival in the cohort compared with the TNM staging. DCA, decision curve analysis; OS, overall survival; TNM, tumor-node-metastasis.

Discussion

Patients with iCCA have poor prognosis because of the aggressiveness of the tumor, delayed clinical diagnosis and lack of effective treatment (26). Despite recent advances in imaging and laboratory tests, the diagnosis of iCCA is still challenging, and patients are often in the late stage, precluding surgical resection (27). Therefore, searching for accurate and effective prognostic factors for iCCA patients will help choose appropriate treatment methods in the early stage. In our study, we found that GLUT1 expression is upregulated in iCCA tissues and correlates with poor prognosis. In addition, we developed a comprehensive prognostic model for evaluating the outcomes of iCCA patients based on clinicopathological features and a novel molecular biomarker GLUT1. This study is the first to use a nomogram based on GLUT1 and clinical features to predict the prognosis of iCCA patients.

GLUT1 is a key rate limiting factor for glucose transport and metabolism in cancer cells (23). In this study, the expression profiling of SLC2A1, by analysis using the TCGA database, was found upregulated in various cancerous tissues, including BRCA, CHOL, COAD and so on, indicating an oncogenic role of SLC2A1 in various types of tumors. A recent functional study has confirmed that overexpression of GLUT1 promoted the proliferation, motility, and invasiveness of iCCA cells in vitro and in vivo (11). And also found GLUT1 can predict the prognosis of large duct type iCCA (12). Here, the data from IHC assay showed that GLUT1 expression is upregulated in iCCA tissues, and GLUT1 expression was significantly correlated with vascular invasion, advanced T stage, advanced N stage and advanced TNM stage. Moreover, iCCA patients with high GLUT1 expression had a worse OS, suggesting that GLUT1 is a promising prognostic predictor and therapeutic target for iCCA patients.

Currently, the means of assessing the prognosis and predicting the progression of iCCA are based on various clinical frameworks, among which are the TNM staging system and pathological differentiation conditions (28). Studies have shown that, with tumor size of iCCA lesions ≤5 cm, 5-year OS of 70–75% can be achieved following surgical resection (27). In the current research, apart from tumor size, older age and vascular invasion were also significant predictors for the prognosis of iCCA following surgical resection (29). Consistent with what current studies have found, here, we found that three factors including age, tumor size and GLUT1 expression were independently correlated with the prognosis of iCCA, using multivariate analysis. These results indicate that various clinical features and biomarkers are involved in the progress of iCCA, suggesting that a comprehensive prognostic model is needed to evaluate the prognosis of patients with iCCA.

Growing evidence indicates that the nomogram offers more benefits than traditional staging systems. This is because it considers multiple predictive factors when presenting prognostic information (30). Moreover, as the nomogram demonstrates good discrimination and precise prognostic ability, it might be helpful in individualized decision-making regarding post-operative treatment and in offering counseling (31). In this study, we established a nomogram based on three independent prognostic factors to evaluate the prognosis of iCCA, which shows that patients older than 60 years would likely have a worse prognosis than younger patients. In addition, this nomogram showed the extent of prognosis deteriorates as tumor size increases. The nomogram also clearly illustrated that patients with higher GLUT1 expression are more likely to die than those with lower GLUT1 expression. Compared with traditional TNM staging systems, this nomogram incorporated the age, tumor size and GLUT1 expression, showing superior discriminatory ability for predicting survival after iCCA resection. Overall, this iCCA nomogram flexibly quantified patients’ risk by using the most crucial predictors. The greater the total score, the poorer the prognosis.

Moreover, certain instructions can be derived for clinicians based on the total score calculated by the nomogram (32). For example, for patients with similar clinical characteristics but different GLUT1 expression, treatment choices may differ based on different life expectancy provided by this nomogram, instead of relying on their clinical experience only. By using the nomogram to distinguish subgroups of iCCA patients with more uniform prognosis, clinicians can more objectively and precisely assess a wide variety of parameters for iCCA. This makes the understanding of clinical trial results clearer. Additionally, this nomogram is useful for evaluating individual clinical outcomes and the potential of specific treatments for iCCA. For example, Consider a 65-year-old male patient with a tumor size of 5 cm and a GLUT1 expression level of 8 (on a scale of 1–10). According to the nomogram, the scores for these characteristics are 40, 30, and 20, respectively. Adding these scores gives a total of 90 points, which corresponds to an estimated 2-year survival rate of 60% and a 3-year survival rate of 40%.

However, we also recognize that this nomogram has certain limitations due to the lack of external validation. Therefore, we plan to collaborate with other institutions in future studies to validate the nomogram in multi-center cohorts to confirm its generalizability. Furthermore, since the prognosis of iCCA involves multiple molecular regulatory pathways, such as genetic mutations IDH1/2, FGFR2 and immune checkpoints programmed death-ligand 1 (PD-L1), our study only focuses on the GLUT1 protein, which limits the comprehensiveness of the nomogram. Future work will integrate multiple biomarkers to further enhance the nomogram, enabling it to more accurately capture the complexity of iCCA prognosis.

Conclusions

This study indicates the expression of GLUT1 in iCCA, providing a therapeutic target and novel molecular biomarker to predict clinical outcomes. Moreover, we established a nomogram incorporating GLUT1 with patient age and tumor size to predict the prognosis of patients with iCCA intuitively and accurately.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-630/rc

Data Sharing Statement: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-630/dss

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-630/prf

Funding: This work was supported by grants from National Natural Science Foundation of China (No. 82203679 to L.P.G.), Medical Research Foundation of Guangdong Province (No. A2022380 to L.P.G. and No. A2023066 to J.L.), Guangdong Basic and Applied Basic Research Foundation (No. 2021A1515111142 to Y.W.Z., No. 2022A1515012555 to C.K.S. and No. 2024A1515012426 to C.K.S.), the Guangzhou Municipal Science and Technology Project (No. 202201011231 to J.L.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-630/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 Institute Research Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University (No. SL-II2024-355-01) and informed consent was obtained from all individual participants.

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. Shaib YH, Davila JA, McGlynn K, et al. Rising incidence of intrahepatic cholangiocarcinoma in the United States: a true increase? J Hepatol 2004;40:472-7. [Crossref] [PubMed]
2. Wu L, Tsilimigras DI, Paredes AZ, et al. Trends in the Incidence, Treatment and Outcomes of Patients with Intrahepatic Cholangiocarcinoma in the USA: Facility Type is Associated with Margin Status, Use of Lymphadenectomy and Overall Survival. World J Surg 2019;43:1777-87. [Crossref] [PubMed]
3. Rahnemai-Azar AA, Weisbrod A, Dillhoff M, et al. Intrahepatic cholangiocarcinoma: Molecular markers for diagnosis and prognosis. Surg Oncol 2017;26:125-37. [Crossref] [PubMed]
4. Esnaola NF, Meyer JE, Karachristos A, et al. Evaluation and management of intrahepatic and extrahepatic cholangiocarcinoma. Cancer 2016;122:1349-69. [Crossref] [PubMed]
5. Li H, Li Y, Kang W, et al. Hepatic arterial infusion chemotherapy versus systemic chemotherapy in unresectable intrahepatic cholangiocarcinoma: a propensity score-matched analysis of efficacy and safety. Quant Imaging Med Surg 2025;15:4387-99. [Crossref] [PubMed]
6. Qin R, Fan X, Huang Y, et al. Role of glucose metabolic reprogramming in colorectal cancer progression and drug resistance. Transl Oncol 2024;50:102156. [Crossref] [PubMed]
7. Wang X, Zhou L, Qi L, et al. High GLUT1 membrane expression and low PSMA membrane expression in Ductal Adenocarcinoma and Intraductal Carcinoma of the prostate. Prostate Cancer Prostatic Dis 2024;27:720-7. [Crossref] [PubMed]
8. Dong F, He K, Zhang S, et al. SSRI antidepressant citalopram reverses the Warburg effect to inhibit hepatocellular carcinoma by directly targeting GLUT1. Cell Rep 2024;43:114818. [Crossref] [PubMed]
9. Huang S, Wang X, Zhu Y, et al. SOX2 promotes vasculogenic mimicry by accelerating glycolysis via the lncRNA AC005392.2-GLUT1 axis in colorectal cancer. Cell Death Dis 2023;14:791. [Crossref] [PubMed]
10. Shao M, Pan Q, Tan H, et al. CYP3A5 unexpectedly regulates glucose metabolism through the AKT-TXNIP-GLUT1 axis in pancreatic cancer. Genes Dis 2024;11:101079. [Crossref] [PubMed]
11. Tiemin P, Peng X, Qingfu L, et al. Dysregulation of the miR-148a-GLUT1 axis promotes the progression and chemoresistance of human intrahepatic cholangiocarcinoma. Oncogenesis 2020;9:19. [Crossref] [PubMed]
12. Kinjo Y, Naito Y, Akiba J, et al. SUOX and GLUT1 are biomarkers for the prognosis in large duct type intrahepatic cholangiocarcinoma. Hum Pathol 2022;128:11-9. [Crossref] [PubMed]
13. Fang C, Wang W, Feng X, et al. Nomogram individually predicts the overall survival of patients with gastroenteropancreatic neuroendocrine neoplasms. Br J Cancer 2017;117:1544-50. [Crossref] [PubMed]
14. Kim Y, Spolverato G, Ejaz A, et al. A nomogram to predict overall survival and disease-free survival after curative resection of gastric adenocarcinoma. Ann Surg Oncol 2015;22:1828-35. [Crossref] [PubMed]
15. Kent MS, Mandrekar SJ, Landreneau R, et al. A Nomogram to Predict Recurrence and Survival of High-Risk Patients Undergoing Sublobar Resection for Lung Cancer: An Analysis of a Multicenter Prospective Study (ACOSOG Z4032). Ann Thorac Surg 2016;102:239-46. [Crossref] [PubMed]
16. Wang Y, Yao J, Wang L, et al. Imaging prediction of lymph node metastasis and early recurrence in patients with distal cholangiocarcinoma. J Gastrointest Oncol 2025;16:1258-67. [Crossref] [PubMed]
17. Mansour JC, Aloia TA, Crane CH, et al. Hilar cholangiocarcinoma: expert consensus statement. HPB (Oxford) 2015;17:691-9. [Crossref] [PubMed]
18. Lu Y, Zheng L, Mei H. Impact of postoperative chemotherapy on the prognosis of combined hepatocellular-cholangiocarcinoma: a retrospective study based on the SEER database. J Gastrointest Oncol 2025;16:1220-32. [Crossref] [PubMed]
19. Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010;17:1471-4.
20. Dong M, Wang HY, Zhao XX, et al. Expression and prognostic roles of PIK3CA, JAK2, PD-L1, and PD-L2 in Epstein-Barr virus-associated gastric carcinoma. Hum Pathol 2016;53:25-34. [Crossref] [PubMed]
21. Sun HW, Yu XJ, Wu WC, et al. GLUT1 and ASCT2 as Predictors for Prognosis of Hepatocellular Carcinoma. PLoS One 2016;11:e0168907. [Crossref] [PubMed]
22. Qin H, Wang Q, Xu J, et al. Integrative analysis of anoikis-related genes prognostic signature with immunotherapy and identification of CDKN3 as a key oncogene in lung adenocarcinoma. Int Immunopharmacol 2024;143:113282. [Crossref] [PubMed]
23. Wright EM. SGLT2 and cancer. Pflugers Arch 2020;472:1407-14. [Crossref] [PubMed]
24. Gao XX, Li JF. Current strategies for predicting post-hepatectomy liver failure and a new ultrasound-based nomogram. World J Gastroenterol 2024;30:4254-9. [Crossref] [PubMed]
25. Wei C, Sun H, Liu S, et al. A nomogram for predicting survival based on hemoglobin A1c and circulating tumor cells in advanced gastric cancer patients receiving immunotherapy. Int Immunopharmacol 2024;142:113239. [Crossref] [PubMed]
26. Macias RIR, Kanzaki H, Berasain C, et al. The Search for Risk, Diagnostic, and Prognostic Biomarkers of Cholangiocarcinoma and Their Biological and Clinicopathologic Significance. Am J Pathol 2025;195:422-36. [Crossref] [PubMed]
27. Sun Y, Jiang W, Duan R, et al. Research progress and prospect of postoperative adjuvant therapy for resectable intrahepatic cholangiocarcinoma. Front Pharmacol 2024;15:1432603. [Crossref] [PubMed]
28. Li Y, Yu J, Zhang Y, et al. Advances in targeted therapy of cholangiocarcinoma. Ann Med 2024;56:2310196. [Crossref] [PubMed]
29. Lv TR, Hu HJ, Ma WJ, et al. Meta-analysis of prognostic factors for overall survival and disease-free survival among resected patients with combined hepatocellular carcinoma and cholangiocarcinoma. Eur J Surg Oncol 2024;50:107279. [Crossref] [PubMed]
30. Sun Y, Huang L, Shen X, et al. Development and validation of a dynamic nomogram for individualized prediction of survival in patients with colon cancer. Sci Rep 2024;14:28033. [Crossref] [PubMed]
31. Zhang Y, Xu W, Wu M, et al. Survival risk stratification based on prognosis nomogram to identify patients with esophageal squamous cell carcinoma who may benefit from postoperative adjuvant therapy. BMC Cancer 2024;24:1330. [Crossref] [PubMed]
32. Jin Z, Zou Q, Zhou T, et al. Preoperative prediction of early recurrence in patients with BRAF mutant colorectal cancer using a intergrated nomogram. Sci Rep 2024;14:25320. [Crossref] [PubMed]