Nomogram-based prognostic stratification for patients with large hepatocellular carcinoma: a population study of SEER database and a Chinese cohort

Introduction

According to the global cancer statistics, hepatocellular carcinoma (HCC) accounts for more than 75% of primary liver cancer, which is the fourth leading cause of cancer-related death worldwide (1). It is reported that the prognosis of patients with HCC is associated with the tumor size (2). Approximately 32% of HCC exhibit a tumor diameter exceeding 5 cm, with an additional 10–20% of tumors greater than 10 cm at the time of diagnosis (3). Guidelines from Asia-Pacific (4), Europe (5) and America (6) regard 5 cm as the cut-off to differentiate early and intermediate stage of HCC. Large HCC, defined as a tumor ≥5 cm in diameter, is associated with a poor survival due to increased potential of invasiveness, metastasis and challenge of surgical resection, raising the need for the evaluation of prognosis for patients with large HCC (7).

Currently, there are several staging systems widely applied to assess the outcome of HCC patients, such as the American Joint Committee on Cancer (AJCC), the Barcelona Clinic Liver Cancer system, the Okuda system. However, these staging systems have limited ability to predict survival for large HCC patients (8). For instance, the AJCC staging system only takes the tumor node metastasis (TNM) classification staging into account, without consideration of additional demographic and clinicopathologic characteristics, which may introduce sampling bias into the prognosis evaluation process (9). Other factors such as physical status of patients, liver function and tumor characteristics can also significantly affect the prognosis of patients (10). Therefore, there is a need for a comprehensive prognostic model that incorporates more significant factors, to assist clinicians in making decisions regarding treatment methods and follow-up plans.

As a graphical calculating model, nomogram has gained increasing interest and application in the diagnosis and prognosis of patients in several cancers such as melanoma, liver cancer and prostate cancer, through simply generating individual survival time or incidence probability by combining all risk factors for disease development (11-13). In the context of HCC, several nomograms have been devised for prognostic purposes; however, the majority of these nomograms were derived from a single population or lack validation with an independent cohort. To our knowledge, a prognostic nomogram for survival prediction in patients with large HCC has not yet been constructed.

This population-based study aimed to establish a reliable nomogram model to predict individual 3- and 5-year overall survival (OS) rates for patients with large HCC-based on a large cohort of population from the Surveillance, Epidemiology, and End Results (SEER) database. Furthermore, the predictive ability of nomogram model was validated compared with the TNM staging system, utilizing data from an external population of patients from our hospital. In addition, we conducted a risk stratification to enable clinicians to select optimal treatment methods and formulate follow-up strategy for patients with large HCC, especially for the high-risk population. We present this article in accordance with the TRIPOD reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-24-288/rc).

Methods

Patients

Data of patients were retrospectively collected from SEER database and the First Affiliated Hospital, Zhejiang University School of Medicine. Information on patients with HCC between 2004 and 2015 were extracted from the SEER database using SEER Stat software (version 8.4.0; NCI, Bethesda, MD, USA) (14). The International Classification of Diseases for Oncology (ICD-O-3 code: 8170, 8171, 8172, 8173, 8174 and 8175) and site codes C22.0 (liver) were used to screen out HCC patients. Inclusion criteria were as follows: (I) HCC as the primary tumor; (II) tumor diameter ≥5 cm; and (III) available demographic characteristics (age, gender, race, insurance and marital status), tumor-related factors [tumor size, histological grade, 6th TNM staging, alpha fetoprotein (AFP), fibrosis score], treatment (surgery, radiation and chemotherapy) and survival information. Exclusion criteria were as follows: (I) age ≤18 years and (II) diagnosis of HCC by death certificate or autopsy. Fibrosis score was scored as 0–4 (none or moderate fibrosis) and 5–6 (severe fibrosis or cirrhosis). There were three categories for treatment including none (no tumor-directed surgery), local tumor destruction (such as photodynamic therapy and cryosurgery) and resection (partial or total hepatectomy).

As for extracting patients from our center, we included 302 patients with large HCC from May 2019 to March 2021 as the external validation cohort. The following inclusion criteria were adopted: (I) aged more than 18 years; (II) diagnosis of HCC with tumor diameter ≥5 cm; (III) no severe chronic diseases; (IV) no other tumor history; and (V) complete data such as tumor size, AFP, treatment, and follow-up information. In cases where tumor resection was not performed, imaging modalities including contrast-enhanced ultrasound, computed tomography (CT), magnetic resonance imaging (MRI) were utilized to assess tumor size and T stage, N stage and M stage. Moreover, the variable of AFP was divided into normal and elevated. OS was computed from the date of diagnosis to death or the last follow-up. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the ethics committee of the First Affiliated Hospital, Zhejiang University School of Medicine (No. 2016-324) and individual consent for this retrospective analysis was waived.

Construction of nomogram-based prognostic stratification

Patients from SEER database were randomly divided into the training cohort and internal validation cohort in a ratio of 7:3. In training cohort, univariate Cox regression analysis was performed to identify the association between OS and variables such as age, gender, tumor size, histological grade, TNM staging, AFP, fibrosis score, surgery and chemotherapy. Subsequently, variables with P value <0.05 in univariate analysis were included in multivariable analysis to determine the independent prognostic indicators. Based on the independent prognostic factors, a nomogram was constructed to predict 3- and 5-year survival rates for patients with large HCC.

The discrimination and calibration of nomogram were assessed by the internal and external validation cohorts compared with TNM staging system. Specifically, the concordance index (C-index) was adopted to evaluate the discriminative ability, though quantifying the discrepancies between observed and predicted outcomes, and higher C-index value meant more accurate prediction (15). Furthermore, calibration plots were applied to reflect the consistency between the actual and predicted survival probabilities. The discrimination of TNM staging system and nomogram was compared via the area under receiver operating characteristic (ROC) curve (AUC). Decision curve analysis (DCA) was used to determine the clinical benefit of nomogram. In addition, the total point of each patient was calculated according to score assignment of variables in the nomogram. Subsequently, a nomogram-based prognostic stratification system was developed to separated patients into high-, middle-, and low-risk groups. Kaplan-Meier (K-M) curves were drawn to estimated visually survival difference among the three risk groups.

Statistical analysis

All statistical analyses were performed using SPSS software (IBM Corporation, USA, version 26) and programming language R (version 3.6.2). Univariable and multivariable Cox regression analyses were conducted to identify the independent predictors associated with OS. Only the variables with P values <0.05 in univariable analysis were subsequently included in multivariable analysis. The cutoff values of the total points from nomogram were determined by X-tile program (Yale University, New Haven, CT, USA). Besides, the survival differences among three groups were estimated by log-rank test. C-index, ROC curves, nomogram, calibration curves, DCA curves and K-M curves were generated in R with packages “rms”, “survival”, “foreign”, “timeROC” and “regplot”. Statistical significance was set as P<0.05 in a two-sided test.

Results

Patient characteristics

A total of 1,426 patients with large HCC from SEER database were eventually included and randomly divided into the training cohort (n=1,012) for nomogram construction and internal validation cohort (n=414) for internal verification. An independent cohort of 302 patients from our center formed the external validation cohort. The detailed process of patient selection is shown in Figure 1. Since the diagnosis of HCC mainly relies on imaging test, unknown histological grade accounted for the majority of HCC. There were more patients treated with local tumor destruction while a smaller proportion received chemotherapy in our center than SEER database. More clinicopathologic characteristics of patients in three cohorts are listed in Table 1.

Figure 1 Flow diagram of the selection criteria of patients with large hepatocellular carcinoma. SEER, Surveillance, Epidemiology, and End Results database; HCC, hepatocellular carcinoma; TNM, tumor node metastasis staging; AFP, alpha fetoprotein.

Independent prognostic factors

Univariate Cox analyses revealed that age, race, histological grade, AJCC stage, T stage, N stage, M stage, AFP, fibrosis score, surgery and chemotherapy were significantly associated with OS in the training set. It is worth mentioning that AJCC stage was not involved in subsequent multivariate analysis due to its dependence from T, N and M stages. Multivariate Cox analysis demonstrated that histological grade, T stage, M stage, AFP, fibrosis score, surgery and chemotherapy were independent prognostic factors of OS (Table 2). As for chemotherapy, there are many different types of chemotherapy with varied therapeutic effects, and it is unclear which chemotherapy drug was used in SEER database. Therefore, chemotherapy is excluded as independent prognostic factor in the following analysis.

Development and validation of nomogram

A nomogram model integrating the independent prognostic factors was established for prediction of 3- and 5-year survival rates for patients with large HCC in the training set (Figure 2). As nomogram showed, surgery made the largest contribution, followed by histological grade, M stage, AFP, fibrosis score and T stage. Each level of the variables is assigned a score on the Points scale in the light of its prognostic value, and the total point can be simply acquired by summing the score of each factor. The sizes of the blue boxes in the middle and the yellow block on the total points axis represented the proportion of patients. The score assignment for variables included in nomogram is shown in Table 3. For instance, as shown in Figure 2, if a patient presented with a fibrosis score of 0–4 (69 points), an elevated AFP level (84 points), histological grade poor/undifferentiated (100 points), T2 stage (80 points) and M1 stage (95 points), and subsequently undergone surgical resection (0 points), his/her total point amounted to 428. This point corresponded to an estimated 3- and 5-year survival rates of 28.0% and 15.8%, respectively. Total points and survival rate in the nomogram were shown in Table 4. Furthermore, an online version of nomogram was provided at https://jikun.shinyapps.io/large_hcc (Figure 3), so as to enable calculation easy and avoid the manual calculation errors.

Figure 2 Nomogram for predicting prognosis of patients with large hepatocellular carcinoma. Each category of the prognostic variables is assigned a score on the points scale. The sum of these scores is located on the total points scale and a line is drawn downward to determine the specific probability of 3- and 5-year overall survival. AFP, alpha fetoprotein.

Figure 3 Layout of an online version of the developed nomogram (https://jikun.shinyapps.io/large_hcc). The black line stands for the estimated survival probability of the patients with a fibrosis score of 0–4, normal AFP level, unknown histological grade, T1 and M0 stage, and no surgery. The blue line stands for those with a fibrosis score of 5–6, normal AFP level, well/moderate histological grade, T1 and M0 stage, and undergoing local tumor destruction. AFP, alpha fetoprotein.

C-indexes of nomogram were 0.730 (95% CI, 0.712–0.748) in the training cohort, 0.726 (95% CI, 0.695–0.757) in the internal validation cohort, and 0.724 (95% CI, 0.677–0.771) in the external validation cohort, manifesting accurate capability in prognosis prediction. The calibration curves for probability of survival at 3 and 5 years exhibited excellent agreement between the actual and predicted observation in three cohorts (Figure 4). Meanwhile, AUCs of nomogram in ROC curves revealed superior discrimination in comparison to that of AJCC TNM staging [training cohort: 3-year survival (0.799 vs. 0.675) and 5-year survival (0.811 vs. 0.676); internal validation cohort: 3-year survival (0.834 vs. 0.683) and 5-year survival (0.873 vs. 0.712); external validation cohort: 3-year survival (0.843 vs. 0.720) and 5-year survival (0.790 vs. 0.685)] (all P<0.01) (Figure 5). Importantly, DCA curves showed that nomogram gained more clinical benefits than TNM staging for predicting OS at 3 and 5 years in three cohorts (Figure 6).

Figure 4 Calibration curves of nomogram for predicting overall survival at 3- and 5-year in the training cohort (A), internal validation cohort (B) and external validation cohort (C), respectively.

Figure 5 ROC curves of nomogram and TNM staging for predicting overall survival at 3- and 5-year in the training cohort (A,B), internal validation cohort (C,D) and external validation cohort (E,F), respectively. ROC, receiver operating characteristic curve; AUC, areas under the curve; TNM, tumor node metastasis staging.

Figure 6 Decision curves of nomogram and TNM staging for clinical utility of overall survival at 3- and 5-year in the training cohort (A,B), internal validation cohort (C,D) and external validation cohort (E,F), respectively. TNM, tumor node metastasis staging.

Construction of prognostic stratification

The median survival time was 19, 17 and 25 months in the training, internal validation and external validation cohorts, respectively. Based on the total points calculated from nomogram, patients were divided into three groups: 345–414 (low-risk group), 415–460 (medium-risk group) and 461–513 (high-risk group). The median OS of patients in low-risk, medium-risk and high-risk group were 45, 14 and 5 months in the training set; 45, 12 and 4 months in the internal validation set; >45, 19 and 8.6 months in the external validation set, respectively. K-M curves of three risk groups were clearly separated in three cohorts (P<0.01), which implied a stronger correlation between lower risk and longer OS (Figure 7).

Figure 7 Kaplan-Meier curves for comparing different risk groups of patients with large hepatocellular carcinoma in the training cohort (A), internal validation cohort (B) and external validation cohort (C).

Discussion

Advances in imaging modalities and health monitoring strategy in high-risk population have facilitated the early detection of HCC, leading to improved outcomes in terms of OS (16). However, large HCC with a diameter ≥5 cm is frequently diagnosed at an advanced stage, presenting a challenge to clinicians because of high likelihood of microvascular invasion, satellite nodules, recurrence and metastasis (17,18). Previous studies have revealed that some factors such as cirrhosis, vascular invasion and treatment are associated with OS in patients with large HCC (9,19). Nevertheless, until now, there is limited study having developed nomogram to predict outcomes for large HCC. In the present study, we developed and validated a nomogram model based on a large population from SEER database and an independent cohort of patients from our center to accurately estimate individual 3- and 5-year survival probabilities for patients with large HCC. The nomogram showed excellent discrimination and calibration of valuable consistencies between the prediction and observation especially in the external validation cohort, indicating a superior external utility of nomogram.

Although the AJCC TNM staging has been widely applied to determine treatment method and expected survival for HCC patients (6), there is still controversy regarding its ability in predicting prognosis (8,12). This nomogram exhibited a superiority over the TNM staging system in predictive ability of prognosis, with a higher AUCs in internal and external validation cohorts. In DCA curves, improved clinical value of nomogram was observed compared with the TNM staging system. These results clearly illustrated considerable difference between the TNM staging based on T, N, and M stages solely and nomogram model integrating various significant clinicopathological factors for prognoses prediction of patients with large HCC.

Liver fibrosis, as a result of chronic liver injury, makes a considerable contribution to the formation of cirrhosis and HCC (20). Accordingly, the current study illustrated a correlation between the worsening prognosis and increasing fibrosis degree, potentially attributed to the advancement of fibrosis leading to liver failure, hepatic encephalopathy, and portal hypertension (21,22). Moreover, the tumor resection rate of patients with none or moderate fibrosis was over twice that of those with severe fibrosis or cirrhosis in this study (64.7% vs. 30.6%, P<0.01), indicating that the severity of fibrosis may impact patient survival by influencing the selection of treatment.

In accordance with findings of Liu et al. (23), the level of AFP turned out to be an independent prognostic factor for large HCC. Two staging system, Cancer of the Liver Italian Program (CLIP) (24) and Chinese University Prognostic Index (CUPI), have emphasized and incorporated the preoperative AFP levels into prognostic assessment. The underlying rationale for the correlation between increased levels of AFP and poorer prognosis remains unclear (25). Studies found that a high level of AFP was related to the dysfunctional antigen-presenting cells and the impaired dendritic cells in patients with HCC (26,27), and this interaction between AFP and the immune system may account for poor outcomes.

It was observed that histological grade, a indicator of biological aggressiveness of HCC, had a stronger correlation with tumor recurrence and survival (12,28). In this study, the median survival of patients with histological grade of well/moderate and poor/undifferentiated were 22 and 9 months, respectively (P<0.01), illustrating that degree of differentiation of tumor posed significant influence on prognosis. Furthermore, our study found that the tumor metastasis rate of poor differentiation was significantly higher than that of well differentiation (19% vs. 9.1%, P<0.01), indicting an increased invasiveness and metastatic nature of poor differentiation. Surgery plays a most important role in impacting prognosis in our study. Although surgical resection remains the mainstay of curative treatment for HCC, only 41% of patients are candidates for hepatectomy because of decompensated liver function, vascular invasion and metastasis (29-31), in accordance with fewer than half of the cases had undergone hepatectomy in this study. When it comes to unresectable HCC, the optional therapy includes radiofrequency ablation (RFA) and transcatheter arterial chemoembolization (TACE) (32).

In addition, prognostic stratification system including high-, middle-, and low-risk groups was constructed based on the total scores from nomogram. As shown in the K-M curves, the population at high risk was significantly associated with poor prognosis. The prognostic stratification system may be helpful to formulate individual therapy method and follow-up plan according to the different risk groups. The patients at low risk could have more treatment options based on liver function and tumor status; for patients at medium risk, a combination of targeted therapy or immunotherapy may be recommended to improve survival; patients at high risk may need palliative treatment. For patients deemed to be at high risk, it is recommended that the follow-up interval should be shortened and adjusted according to tumor status. Conversely, for patients classified as low risk group, the follow-up interval may be prolonged appropriately.

There are some limitations in the present study. First, the retrospective design of the study may lead to unavoidable selection bias. Second, nomogram did not include prognostic predictors such as liver function, since information of liver function was unavailable in SEER database. Besides, another limitation was a lack of validation using data of patients from other countries, though we have confirmed the reliable predictive power of nomogram-based prognostic stratification in Chinese patients. Therefore, further multicenter studies should be warranted in order to verify the prognostic stratification worldwide.

Conclusions

In summary, this study proposed a nomogram model on basis of a large population to predict 3- and 5-year survival rates for patients with large HCC, which showed preferable prediction ability in comparison with the TNM staging. Furthermore, nomogram-based prognostic stratification was developed as a valuable tool to assist clinicians on the selection of optimal treatment method and follow-up plan, particularly for high-risk population.

Acknowledgments

We sincerely thank the Surveillance, Epidemiology, and End Results (SEER) program for the efforts in providing high-quality open resources.

Funding: This study was supported by the National Natural Science Foundation of China (Nos. 82102159, 82102160, 12090020, and 12090025), and Zhejiang Provincial Natural Science Foundation Committee-Zhejiang Society for Mathematical Medicine Joint Fund Major Project (No. LSD19H180003).

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-24-288/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-24-288/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-24-288/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-24-288/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 (as revised in 2013). The study was approved by the ethics committee of the First Affiliated Hospital, Zhejiang University School of Medicine (No. 2016-324) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Qi J, Li M, Wang L, et al. National and subnational trends in cancer burden in China, 2005-20: an analysis of national mortality surveillance data. Lancet Public Health 2023;8:e943-55. [Crossref] [PubMed]
2. Ji K, Zhu H, Wu W, et al. Tumor Response and Nomogram-Based Prognostic Stratification for Hepatocellular Carcinoma After Drug-Eluting Beads Transarterial Chemoembolization. J Hepatocell Carcinoma 2022;9:537-51. [Crossref] [PubMed]
3. Chakraborty E, Sarkar D. Emerging Therapies for Hepatocellular Carcinoma (HCC). Cancers (Basel) 2022;14:2798. [Crossref] [PubMed]
4. Hwang SH, Park MS, Park S, et al. Comparison of the current guidelines for diagnosing hepatocellular carcinoma using gadoxetic acid-enhanced magnetic resonance imaging. Eur Radiol 2021;31:4492-503. [Crossref] [PubMed]
5. Italian Association for the Study of the Liver (AISF) , AISF Expert Panel , AISF Coordinating Committee et al. Position paper of the Italian Association for the Study of the Liver (AISF): the multidisciplinary clinical approach to hepatocellular carcinoma. Dig Liver Dis 2013;45:712-23. [Crossref] [PubMed]
6. Chun YS, Pawlik TM, Vauthey JN. 8th Edition of the AJCC Cancer Staging Manual: Pancreas and Hepatobiliary Cancers. Ann Surg Oncol 2018;25:845-7.
7. Arvind A, Memel ZN, Philpotts LL, et al. Thiazolidinediones, alpha-glucosidase inhibitors, meglitinides, sulfonylureas, and hepatocellular carcinoma risk: A meta-analysis. Metabolism 2021;120:154780. [Crossref] [PubMed]
8. He Z, She X, Liu Z, et al. Advances in post-operative prognostic models for hepatocellular carcinoma. J Zhejiang Univ Sci B 2023;24:191-206. [Crossref] [PubMed]
9. Tzedakis S, Sebai A, Jeddou H, et al. Resection Postradioembolization in Patients With Single Large Hepatocellular Carcinoma. Ann Surg 2023;278:756-62. [PubMed]
10. Liu R, Qian MP, Cui YY. Protein kinases: The key contributors in pathogenesis and treatment of nonalcoholic fatty liver disease-derived hepatocellular carcinoma. Metabolism 2023;147:155665. [Crossref] [PubMed]
11. Maurichi A, Miceli R, Eriksson H, et al. Factors Affecting Sentinel Node Metastasis in Thin (T1) Cutaneous Melanomas: Development and External Validation of a Predictive Nomogram. J Clin Oncol 2020;38:1591-601. [Crossref] [PubMed]
12. Chen SH, Wan QS, Zhou D, et al. A Simple-to-Use Nomogram for Predicting the Survival of Early Hepatocellular Carcinoma Patients. Front Oncol 2019;9:584. [Crossref] [PubMed]
13. Wang X, Zhao M, Zhang C, et al. Establishment and Clinical Application of the Nomogram Related to Risk or Prognosis of Hepatocellular Carcinoma: A Review. J Hepatocell Carcinoma 2023;10:1389-98. [Crossref] [PubMed]
14. Che WQ, Li YJ, Tsang CK, et al. How to use the Surveillance, Epidemiology, and End Results (SEER) data: research design and methodology. Mil Med Res 2023;10:50. [Crossref] [PubMed]
15. Longato E, Vettoretti M, Di Camillo B. A practical perspective on the concordance index for the evaluation and selection of prognostic time-to-event models. J Biomed Inform 2020;108:103496. [Crossref] [PubMed]
16. Wang Y, Li M, Zhang Z, et al. Application of Radiomics in the Efficacy Evaluation of Transarterial Chemoembolization for Hepatocellular Carcinoma: A Systematic Review and Meta-analysis. Acad Radiol 2024;31:273-85. [Crossref] [PubMed]
17. Ji K, Shi Y, Liang Z, et al. Lipiodol combined with drug-eluting beads versus drug-eluting beads alone for transarterial chemoembolization of hepatocellular carcinoma: a multicenter study. Acad Radiol 2024; Epub ahead of print. [Crossref] [PubMed]
18. Nguyen-Khac V, Brustia R, Rhaiem R, et al. Liver resection for single large hepatocellular carcinoma: a prognostic factors study. Ann Hepatol 2022;27:100739. [Crossref] [PubMed]
19. Tang L, Hou Y, Huang Z, et al. Transarterial Chemoembolization Combined with Lenvatinib Plus Sequential Microwave Ablation for Large Hepatocellular Carcinoma Beyond Up-to-Seven Criteria: A Retrospective Cohort Study. Acad Radiol 2024;31:2795-806. [Crossref] [PubMed]
20. Kisseleva T, Brenner D. Molecular and cellular mechanisms of liver fibrosis and its regression. Nat Rev Gastroenterol Hepatol 2021;18:151-66. [Crossref] [PubMed]
21. Ginès P, Krag A, Abraldes JG, et al. Liver cirrhosis. Lancet 2021;398:1359-76. [Crossref] [PubMed]
22. Guixé-Muntet S, Zhu CP, Xie WF, et al. Novel therapeutics for portal hypertension and fibrosis in chronic liver disease. Pharmacol Ther 2020;215:107626. [Crossref] [PubMed]
23. Liu PH, Su CW, Hsu CY, et al. Solitary Large Hepatocellular Carcinoma: Staging and Treatment Strategy. PLoS One 2016;11:e0155588. [Crossref] [PubMed]
24. Chan AW, Chong CC, Mo FK, et al. Incorporating albumin-bilirubin grade into the cancer of the liver Italian program system for hepatocellular carcinoma. J Gastroenterol Hepatol 2017;32:221-8. [Crossref] [PubMed]
25. Ishteyaque S, Singh G, Yadav KS, et al. Cooperative STAT3-NFkB signaling modulates mitochondrial dysfunction and metabolic profiling in hepatocellular carcinoma. Metabolism 2024;152:155771. [Crossref] [PubMed]
26. Munson PV, Adamik J, Hartmann FJ, et al. Polyunsaturated Fatty Acid-Bound α-Fetoprotein Promotes Immune Suppression by Altering Human Dendritic Cell Metabolism. Cancer Res 2023;83:1543-57. [Crossref] [PubMed]
27. Munson PV, Adamik J, Butterfield LH. Immunomodulatory impact of α-fetoprotein. Trends Immunol 2022;43:438-48. [Crossref] [PubMed]
28. Wu C, Du X, Zhang Y, et al. Five machine learning-based radiomics models for preoperative prediction of histological grade in hepatocellular carcinoma. J Cancer Res Clin Oncol 2023;149:15103-12. [Crossref] [PubMed]
29. Chen Z, Xie H, Hu M, et al. Recent progress in treatment of hepatocellular carcinoma. Am J Cancer Res 2020;10:2993-3036. [PubMed]
30. Lin WD, Ye LN, Song ZS, et al. Wide surgical margins improve prognosis for HCC with microvascular invasion. Eur Rev Med Pharmacol Sci 2023;27:2052-9. [PubMed]
31. Foerster F, Galle PR. The Current Landscape of Clinical Trials for Systemic Treatment of HCC. Cancers (Basel) 2021;13:1962. [Crossref] [PubMed]
32. Hu H, Chen GF, Yuan W, et al. Microwave ablation with chemoembolization for large hepatocellular carcinoma in patients with cirrhosis. Int J Hyperthermia 2018;34:1351-8. [Crossref] [PubMed]