Li GC, Yao ZY, Mao HS, Han ZX. Association of triglyceride-glucose index with long-term prognosis in advanced hepatocellular carcinoma patients receiving immunotherapy and targeted therapy. World J Gastroenterol 2025; 31(30): 109863 [DOI: 10.3748/wjg.v31.i30.109863]
Corresponding Author of This Article
Zheng-Xiang Han, PhD, Professor, Department of Oncology, The Affiliated Hospital of Xuzhou Medical University, No. 99 Huaihai West Road, Xuzhou 221000, Jiangsu Province, China. xzpxlgc@163.com
Research Domain of This Article
Oncology
Article-Type of This Article
Retrospective Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
World J Gastroenterol. Aug 14, 2025; 31(30): 109863 Published online Aug 14, 2025. doi: 10.3748/wjg.v31.i30.109863
Association of triglyceride-glucose index with long-term prognosis in advanced hepatocellular carcinoma patients receiving immunotherapy and targeted therapy
Geng-Chen Li, Zhi-Yuan Yao, Zheng-Xiang Han, Department of Oncology, The Affiliated Hospital of Xuzhou Medical University, Xuzhou 221000, Jiangsu Province, China
Hong-Sen Mao, Department of Oncology, People’s Hospital of Jiawang District of Xuzhou City, Xuzhou 221000, Jiangsu Province, China
Co-corresponding authors: Hong-Sen Mao and Zheng-Xiang Han.
Author contributions: Li GC and Yao ZY were responsible for the methodology of this study, and took part in the data curation of this study; Li GC, Yao ZY, Mao HS, and Han ZX contributed to the conceptualization, writing-review and editing of this manuscript; Li GC contributed to the formal analysis of this manuscript and the visualization of this article; Li GC, Yao ZY, and Mao HS took part in the writing-original draft and investigation of this manuscript; Li GC, Yao ZY, and Han ZX contributed to the project administration and the supervision of this manuscript; Yao ZY and Mao HS were responsible for the validation of this manuscript; Li GC and Han ZX took part in the resources; Mao HS and Han ZX were involved in the supervision of this study. Li GC and Yao ZY contributed equally to this manuscript, they are co-first authors of this manuscript. Han ZX and Mao HS contributed equally to this manuscript, they are co-corresponding authors of this study.
Institutional review board statement: This retrospective study was approved by the Ethics Review Committee of the Affiliated Hospital of Xuzhou Medical University (Approval No. XYFY2022-KL481-01) and adhered to the principles outlined in the Declaration of Helsinki.
Informed consent statement: Given the retrospective design of this investigation, the Ethics Committee of the Affiliated Hospital of Xuzhou Medical University granted us an exemption from obtaining written informed consent.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: The datasets used and analysed during the current study available from the corresponding author on reasonable request.
Open Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Zheng-Xiang Han, PhD, Professor, Department of Oncology, The Affiliated Hospital of Xuzhou Medical University, No. 99 Huaihai West Road, Xuzhou 221000, Jiangsu Province, China. xzpxlgc@163.com
Received: May 26, 2025 Revised: June 4, 2025 Accepted: July 17, 2025 Published online: August 14, 2025 Processing time: 75 Days and 18.5 Hours
Abstract
BACKGROUND
Primary liver cancer, particularly hepatocellular carcinoma (HCC), ranks as the sixth most prevalent cancer globally and the third major cause of cancer-associated mortality. Despite the available immunotherapies and combined immunotherapy and targeted therapy, the prognosis for many patients remains dismal. Accurately identifying the appropriate patient cohorts is crucial for improving treatment outcomes.
AIM
To investigate the prognostic value of the triglyceride-glucose (TyG) index - a novel, accessible marker of insulin resistance - in predicting therapeutic outcomes among patients with hepatitis B virus (HBV)-related HCC treated with camrelizumab and lenvatinib.
METHODS
In this study, we conducted a retrospective review of 278 patients diagnosed with stage B/C HBV-related HCC who underwent combination therapy. Based on their TyG index, patients were categorized into high and low TyG index groups. A nomogram prediction model was developed based on independent prognostic factors for overall survival (OS) and validated using the C-index and calibration curves.
RESULTS
Of the 278 patients enrolled in the study, 144 were assigned to the high TyG index group, while the remaining 134 were classified into the low index group. Importantly, patients with a low TyG index demonstrated a significantly prolonged median progression-free survival and OS relative to those with a high index. Additionally, the objective response rate and disease control rate were 22.39% and 64.18% in the low TyG index group, whereas they were 12.50% and 51.39% in the high TyG index group, respectively. Moreover, the incidence of hypertension was higher in the high TyG index group than in the low TyG index group. The incidence of other adverse effects did not differ significantly between the groups. Multivariate regression analysis identified independent prognostic factors for OS, including the Barcelona Clinic Liver Cancer stage, alpha-fetoprotein level, Eastern Cooperative Oncology Group score, distant metastasis, and the TyG index. The risk ratio of the TyG index was 0.48 (95% confidence interval: 0.31-0.72, P < 0.001).
CONCLUSION
The TyG index is a reliable long-term predictor of response to combined immunotherapy and targeted therapy in patients with HBV-related HCC. Patients with a low TyG index tend to experience better clinical outcomes.
Core Tip: Primary liver cancer ranks among the most prevalent and lethal malignancies worldwide. Although combination therapies have improved treatment options, overall prognosis remains unsatisfactory. The triglyceride-glucose index, a simple surrogate marker of insulin resistance, has recently been associated with tumor progression and immune modulation. This retrospective study evaluates the prognostic significance of the triglyceride-glucose index in patients with advanced hepatocellular carcinoma receiving immunotherapy combined with targeted therapy, offering insights into its potential role in risk stratification and clinical decision-making.
Citation: Li GC, Yao ZY, Mao HS, Han ZX. Association of triglyceride-glucose index with long-term prognosis in advanced hepatocellular carcinoma patients receiving immunotherapy and targeted therapy. World J Gastroenterol 2025; 31(30): 109863
Hepatocellular carcinoma (HCC), the primary liver cancer type, ranks sixth in global cancer prevalence[1]. Most HCC cases in China are closely associated with hepatitis B virus (HBV) infection[2]. Despite advancements in HCC treatment, many patients still encounter issues such as limited treatment choices and poor prognosis. This is mainly because HCC is often diagnosed at an advanced stage, causing patients to miss potential optimal treatment opportunities[3]. Hence, non-surgical approaches, such as immunotherapy and targeted therapy, are paramount in treating HCC. Numerous research studies and comprehensive meta-analyses have consistently demonstrated that patients with HBV-associated HCC have significantly better outcomes after receiving a combination of immunotherapy and targeted therapy than those with HCC unrelated to viral causes[4,5].
Lenvatinib is an oral tyrosine kinase inhibitor that acts on multiple targets and possesses anti-tumor and anti-angiogenic properties. It is mainly utilized to decelerate the growth rate of tumors and delay disease progression[6]. Nevertheless, studies have revealed that the effect of lenvatinib treatment in some patients with advanced HCC is limited[7]. Camrelizumab is an immune checkpoint inhibitor that relieves immune system suppression and activates T cells to enhance their ability to attack tumors[8]. Finn et al[9] showed that the combination of these drugs significantly improved the treatment effect in patients with advanced HCC. In advanced HCC, the dual approach of immunotherapy and targeted therapies effectively suppresses tumor progression while boosting immune function, resulting in prolonged patient survival and better quality of life[10]. Consequently, this combination therapy has become an important option for patients with advanced liver cancer.
The triglyceride-glucose (TyG) index, calculated by multiplying serum triglyceride and glucose levels, is a widely recognized marker of metabolic health and serves as a reliable indicator of insulin resistance[11]. Recent studies have also suggested its potential as a biomarker for type 2 diabetes, colorectal cancer, and breast cancer[12,13]. Insulin resistance has been associated with the prognosis of advanced HCC as it promotes tumor growth and metastasis by creating a pro-inflammatory environment, altering the immune responses, and potentially impairing immune cell function. These factors contribute to increased resistance to immunotherapy[14]. Martini et al[15] suggested that the TyG index could predict improved prostate cancer survival due to its association with enhanced programmed cell death protein-1 (PD-1) expression. Obesity-associated immune suppression through the enhanced sensitivity to PD-1 inhibitors might further contribute to this enhancement. However, the long-term prognostic value of the TyG index in patients with advanced HCC receiving combined immunotherapy and targeted therapy has not been thoroughly investigated. Filling this knowledge gap is important, as insulin resistance influences tumor progression and immune responses, both of which are key to immunotherapy effectiveness. This retrospective study aimed to bridge this gap by examining whether the TyG index could be a standalone prognostic predictor for patients with advanced HCC.
MATERIALS AND METHODS
Study design and patients
This study retrospectively analyzed the clinical data of patients with liver cancer admitted to the Affiliated Hospital of Xuzhou Medical University between December 2020 and December 2023. The inclusion criteria were: (1) A definitive diagnosis of advanced liver cancer confirmed through pathology and imaging; (2) Combined treatment with camrelizumab and lenvatinib; (3) Available biochemical blood indicators before treatment; (4) Completion of a minimum of two treatment cycles; (5) At least one measurable tumor lesion present; (6) Aged between 18 and 75, with an Eastern Cooperative Oncology Group (ECOG) performance status of 2 or less and an anticipated survival period of at least two months; (7) Child-Pugh classification grade A or B; (8) Barcelona Clinic Liver Cancer (BCLC) stage B or C; and (9) Serum hepatitis B surface antigen positive or HBV-DNA positive.
Patients were excluded if they had: (1) Prior treatment with PD-1 inhibitors or targeted therapies; (2) Significant infections or widespread inflammation; (3) Autoimmune diseases; (4) Long-term use of corticosteroids or immunosuppressants; (5) Severe organ dysfunction of the heart, liver, kidneys, and other major organs; (6) Other concurrent malignancies or non-primary HCC; (7) Incomplete or inaccurate clinical data records; and (8) HBV reactivation during the combined treatment. Finally, 278 participants were analyzed. The patient screening flow chart is shown in Figure 1. The Ethics Committee of the Affiliated Hospital of Xuzhou Medical University approved this study (Approval No. XYFY2022-KL481-01).
Figure 1 Flow chart depicting the screening of hepatocellular cancer patients who received camrelizumab combined with lenvatinib.
PD-1: Programmed cell death protein-1; PD-L1: Programmed death ligand-1; TyG: Triglyceride-glucose.
Definitions
The TyG index was calculated using the formula: TyG = ln(fasting triglyceride × fasting glucose/2), where “ln” denotes the natural logarithm. All biochemical measurements were conducted in mg/dL units. Fasting glucose and triglyceride levels were determined using enzymatic methods with an automated biochemical analyzer (AU2700, Olympus). Standard laboratory protocols were strictly followed to ensure consistency across participants. A receiver operating characteristic curve was constructed for the TyG index before administering the combined treatment. The prognostic performance of the TyG index was evaluated using the area under the curve. The optimal threshold for the TyG index (1.58) was determined using the maximum product method of sensitivity and specificity. Based on this cutoff value, the patients were assigned to the high (≥ 1.58) or low (< 1.58) TyG index group.
Grouping and treatment protocol
Camrelizumab (200 mg) was administered intravenously once every three weeks. Lenvatinib was administered as 4-mg capsules at 8 mg/day to patients < 60 kg and 12 mg/day for patients ≥ 60 kg.
Evaluation
Enhanced computed tomography or magnetic resonance imaging scans were reviewed every 6-8 weeks during treatment to assess the efficacy of treating the target lesion. To ensure accuracy, all imaging data were evaluated by two independent radiologists with more than five years of clinical experience. The long-term treatment effect was evaluated using overall survival (OS) and progression-free survival (PFS). Short-term effectiveness was evaluated using mRECIST criteria. Adverse events (AEs) were documented and analyzed according to the National Cancer Institute’s Common Terminology Criteria for Adverse Events (version 5.0).
Statistical analysis
Statistical analysis was performed using R software, version 4.4.0. Continuous variables are reported as means and standard deviations, while categorical variables are presented as counts and percentages. Group pairs were compared using the Fisher’s exact test, χ2 test, or t-test, as suitable. The median OS and PFS were assessed using the Kaplan-Meier approach and compared using Cox proportional hazards regression. Independent predictors of OS and PFS were identified using multivariate analysis, focusing on significant variables at P < 0.05 in the univariate analysis. The independent predictors for OS included the TyG index, BCLC grade, ECOG score, the presence of distant metastasis, and alpha-fetoprotein (AFP) levels. A nomogram prediction model was developed to predict outcomes at 12, 15, and 18 months. For this nomogram, we selected variables based on their statistical and clinical importance and assigned weights according to their hazard ratios from the multivariate analysis. The model’s internal validation was achieved through a bootstrap technique with 1000 resamples, ensuring the precision of the predictions. All statistical comparisons were two-tailed, and statistical significance was set at P < 0.05.
RESULTS
Patient characteristics
This retrospective analysis examined 278 individuals diagnosed with advanced HCC who underwent treatment with a combination of camrelizumab and lenvatinib from December 2020 to December 2023 (Figure 1). Of these patients, 134 fell into the low TyG index group, and 144 were categorized into the high TyG index group. We excluded 374 patients because of missing clinical data (n = 136), combination with other malignancies or non-primary HCC (n = 32), loss to follow-up for more than six months (n = 75), use of other immunosuppressant or antiangiogenic agents (n = 90), and missing pre-treatment TyG index data (n = 41). Table 1 presents patient baseline data, including clinicopathological characteristics such as sex, age, body mass index (BMI), smoking and drinking status, ECOG score, hypertension status, diabetes status, extrahepatic metastasis, Child-Pugh score, BCLC stage, and the use of interventional treatments, and various laboratory parameters, such as fasting blood glucose, fasting triglycerides, AFP, and HBV-DNA quantification. Baseline BMI, the TyG index, fasting blood glucose, and fasting triglycerides differed significantly between the two groups (all P < 0.001).
Table 1 Baseline characteristics of triglyceride-glucose index < 1.58 and triglyceride-glucose index ≥ 1.58, mean ± SD/n (%).
Table 2 shows the tumor response. A complete response was observed in one patient, while 47 patients had a partial response, 112 had stable disease, and 118 developed progressive disease. The objective response rate was 12.5% for the high TyG index group and 22.4% for the low TyG index group (P = 0.031). The disease control rate was 51.4% in the high TyG index group and 64.2% in the low TyG index group (P = 0.029).
Table 2 Tumor responses of triglyceride-glucose index < 1.58 and triglyceride-glucose index ≥ 1.58, n (%).
The median PFS in the low TyG index group (10.60; range 9.62-12.30 months) was significantly higher than in the high TyG index group [7.65; range 7.10-8.28 months; P < 0.001; hazard ratio (HR) = 0.381; 95% confidence interval (CI): 0.275-0.528; Figure 2A]. Likewise, the median OS in the low TyG index cohort (22.30; range 21.15-23.70 months) was higher than in the high TyG index group (15.71; 13.73-17.65 months; P < 0.001; HR = 0.204; 95%CI: 0.139-0.301; Figure 2B).
Figure 2 Effects of different triglyceride-glucose indexes on the long-term prognosis of hepatocellular cancer patients.
A: Kaplan-Meier plot of in the triglyceride-glucose (TyG) index < 1.58 and TyG index ≥ 1.58 groups; B: Kaplan-Meier plot of overall survival in the TyG index < 1.58 and TyG index ≥ 1.58 groups. TyG: Triglyceride-glucose; HR: Hazard ratio; CI: Confidence interval.
Univariate and multivariate Cox proportional hazards regression analyses
Table 3 presents findings from univariate and multivariate Cox proportional hazards regression assessments of PFS. The key predictors of PFS identified in the multivariate model included ECOG performance status (HR = 1.68; 95%CI: 1.11-2.53; P = 0.014), TyG index (HR = 0.55; 95%CI: 0.37-0.83; P = 0.005), BCLC stage (HR = 1.97; 95%CI: 1.32-2.94; P < 0.001), extrahepatic metastasis (HR = 1.66; 95%CI: 1.13-2.45; P = 0.010), AFP level (HR = 1.73; 95%CI: 1.15-2.62; P = 0.009), and BMI (HR = 0.55; 95%CI: 0.31-0.96; P = 0.037). Similarly, Table 4 shows the results of univariate and multivariate Cox proportional hazards regression analyses of OS. Independent predictors of OS included BCLC stage (HR = 1.87; 95%CI: 1.25-2.80; P = 0.002), AFP level (HR = 1.51; 95%CI: 1.01-2.28; P = 0.046), extrahepatic metastasis (HR = 1.58; 95%CI: 1.07-2.33; P = 0.021), TyG index (HR = 0.48; 95%CI: 0.31-0.72; P < 0.001), and ECOG performance status (HR = 1.71; 95%CI: 1.12-2.61; P = 0.012). We used these independent OS predictors to develop nomograms to predict patient survival at 12, 15, and 18 months (Figure 3).
Patients were randomly divided into a training set (n = 194) and an internal validation set (n = 84) at a 7:3 ratio. The baseline characteristics of the two groups are summarized in Table 5. The corrected prediction model had a C-index of 0.81 (95%CI: 0.75-0.86) in the training set and 0.78 (95%CI: 0.73-0.82) in the validation set, indicating that the model had good prediction accuracy. As illustrated in Figure 4A, the area under the curve for a 12-month OS was 0.80 (95%CI: 0.73-0.87) in the training set and 0.87 (95%CI: 0.78-0.96) in the internal validation set. The respective values were 0.80 (95%CI: 0.73-0.86) and 0.93 (95%CI: 0.87-0.99) for 15-month OS (Figure 4B), and 0.85 (95%CI: 0.79-0.91) and 0.90 (95%CI: 0.82-0.98) for 18-month OS (Figure 4C). The OS calibration curves for the training and validation sets at 12, 15, and 18 months (Figure 5) indicated that the predicted risk closely matched the observed risk.
Figure 4 Graph depicting the operating characteristic evaluation plot for a prognostic model.
A: Graph showing the training set and validation set receiver operating characteristic (ROC) evaluation plots for 12-month prognostic prediction model; B: Graph showing the training set and validation set ROC evaluation plots for 15-month prognostic prediction model; C: Graph showing the training set and validation set ROC evaluation plots for 18-month prognostic prediction model. AUC: Area under the curve; CI: Confidence interval.
Figure 5 Graph illustrating the calibration plots for a prognostic model.
A: Calibration plots for the training set 12-month overall survival (OS); B: Calibration plots for the validation set 12-month OS; C: Calibration plots for the training set 15-month OS; D: Calibration plots for the validation set 15-month OS; E: Calibration plots for the training set 18-month OS; F: Calibration plots for validation set 18-month OS.
Table 5 Comparison of features between the training and validation sets, n (%).
Characteristic
Test (n = 84)
Train (n = 194)
χ2
P value
Age, years
0.10
0.747
< 60
39 (46.43)
86 (44.33)
≥ 60
45 (53.57)
108 (55.67)
Sex
0.79
0.375
Female
46 (54.76)
95 (48.97)
Male
38 (45.24)
99 (51.03)
BMI, kg/m2
0.81
0.369
< 25
64 (76.19)
157 (80.39)
≥ 25
20 (23.81)
37 (19.07)
TyG index
0.02
0.894
Low
41 (48.81)
93 (47.94)
High
43 (51.19)
101 (47.94)
HBV-DNA
0.80
0.371
HBV ≤ 2000
41 (48.81)
106 (54.64)
HBV > 2000
43 (51.19)
88 (45.36)
HBe
0.08
0.778
Negative
62 (73.81)
140 (72.16)
Positive
22 (26.19)
54 (27.84)
ECOG
0.07
0.785
0-1
54 (64.29)
128 (65.98)
2
30 (35.71)
85 (34.02)
Child-Pugh score
0.00
0.961
B
44 (52.38)
101 (56.19)
C
40 (47.62)
93 (43.81)
PVTT
1.55
0.213
No
60 (71.43)
152 (78.35)
Yes
24 (28.57)
42 (21.65)
Diabetes
2.41
0.121
No
59 (70.24)
153 (83.15)
Yes
25 (29.76)
41 (16.85)
AFP, ng/mL
0.65
0.420
< 1210
35 (41.67)
91 (46.91)
≥ 1210
49 (58.33)
103 (53.09)
BCLC class
1.58
0.208
B
39 (46.43)
106 (54.64)
C
45 (53.57)
88 (45.36)
Metastasis
1.66
0.197
No
62 (73.81)
128 (65.98)
Yes
22 (26.19)
66 (34.02)
ALT
0.76
0.384
≤ 40
42 (50.00)
108 (55.67)
> 40
42 (50.00)
86 (44.33)
Total bilirubin
0.22
0.642
≤ 34
54 (64.29)
119 (61.34)
> 34
30 (35.71)
75 (38.66)
Cirrhosis
0.07
0.798
No
46 (54.76)
103 (53.09)
Yes
38 (45.24)
91 (46.91)
Interventional
0.57
0.451
No
50 (59.52)
106 (54.64)
Yes
34 (40.48)
88 (45.36)
Hypertensive
0.13
0.720
No
46 (54.76)
123 (63.40)
Yes
38 (45.24)
50 (36.60)
AEs
AEs of varying severity were reported. The most frequently observed AEs included rashes, nausea, fatigue, diarrhea, hypothyroidism, hypertension, and reactive cutaneous capillary endothelial proliferation (Table 6). None of the patients discontinued treatment due to AEs. Patients in the high TyG index group were more likely to experience hypertension than those in the low TyG index group (P < 0.05), while the incidence of the other AEs was similar between the groups (P > 0.05).
Table 6 Treatment-related adverse events in patients with hepatocellular camrelizumab, n (%).
This research marks the first in-depth investigation into the relationship between the TyG index and the effectiveness of combined immunotherapy and targeted therapy in patients with HBV-related HCC. The findings revealed that individuals with a low TyG index experienced substantially greater benefits from the combined treatment in terms of short-term treatment outcomes and long-term survival than those with a higher TyG index. Key independent predictors of PFS included the ECOG score, BMI, AFP level, BCLC stage, distant metastasis, and the TyG index. OS was independently associated with the ECOG score, AFP level, BCLC stage, distant metastasis, and the TyG index. Notably, patients with a high TyG index had a higher likelihood of developing hypertension (P < 0.05). No significant differences were observed in other AEs between the groups (P > 0.05), and death was not associated with AEs.
With the advent of the immunotherapy era, the treatment regimens for patients with HBV-related HCC, including immunotherapy and combination therapy, have undergone considerable optimization[16,17]. However, there is still a lack of relevant studies predicting the long-term prognosis of such patients. In recent years, research on the TyG index has generated increasing attention, but most studies analyzed patients with cardiovascular diseases or metabolic imbalances[18,19]. This study was the first to introduce the TyG index into combined immunotherapy and targeted therapy for liver cancer. Unlike the limited predictive value of the TyG index observed in previous cancer studies[20], the present study found that a low TyG index predicted a good prognosis for patients with HBV-related HCC receiving combined immunotherapy and targeted therapy. This result might be due to differences in tumor types and immunotherapy regimens. T-cell activation relies on the dynamic balance between glucose metabolism (glycolysis) and oxidative phosphorylation[21]. The competitive consumption of glucose by tumor and other immunosuppressive cells (such as regulatory T cells) due to insulin resistance and a hyperglycemic state in patients with a high TyG index limits the acquisition of sufficient energy by effector T cells (such as CD8+ T cells) and suppresses their anti-tumor activity[22,23]. This energy deficiency could induce a metabolic stress state that promotes the inhibition of receptors such as PD-1 and lymphocyte activation gene 3, contributing to immune exhaustion and facilitating tumor immune escape[24,25]. Moreover, insulin resistance and hyperglycemia have been shown to enhance the suppressive activity of regulatory T cells and myeloid-derived suppressor cells, further disrupting anti-tumor immune responses[26]. While our proposed mechanisms linking the TyG index to immunotherapy efficacy are supported by existing literature, we acknowledge that these remain hypothetical due to the lack of direct biological data in our study. Future research incorporating tumor tissue analyses, immune cell profiling, and inflammatory cytokine measurements will be necessary to validate the metabolic-immunologic pathways suggested here. In the context of HBV-related HCC, this effect might be further exacerbated by chronic liver inflammation and viral-induced immunosuppression. Effector T cells struggle to access metabolic substrates due to glucose competition and are functionally impaired by the insulin resistance-induced activation of the phosphatidylinositol 3-kinase/protein kinase B/mammalian target of the rapamycin (PI3K/Akt/mTOR) signaling pathway[27]. This results in a dual hit - metabolic exhaustion and upregulation of immune checkpoint proteins like PD-1 - that contributes to T cell dysfunction and immune escape[28]. Such mechanisms provide a plausible explanation for the reduced efficacy of PD-1 blockade in patients with a high TyG index, as observed in our study. Additionally, a high cholesterol state can weaken the functions of effector T cells and natural killer cells by activating immunosuppressive pathways in the tumor microenvironment[29]. Patients with a low TyG index typically have lower cholesterol levels, possibly resulting in reduced expression of immunosuppressive factors in the tumor microenvironment. Furthermore, the chronic inflammation associated with a high TyG index cannot be ignored. A high TyG index often reflects higher insulin resistance, more severe fat metabolism disorders, and poorer glucose metabolism status than a low TyG index. These are usually accompanied by elevated systemic inflammation[30,31]. The chronic inflammation in patients with HBV-related HCC is jointly driven by viral infection, liver fibrosis, and metabolic disorders[32]. This high inflammatory burden leads to high levels of pro-inflammatory factors (such as interleukin-6 and tumor necrosis factor-α) in the tumor microenvironment[33], which might further limit the recovery of immune cells and anti-angiogenesis processes[34,35]. Studies have demonstrated that insulin resistance is closely related to the excessive activation of the PI3K/Akt/mTOR signaling pathway[36]. The abnormal activity of this pathway could promote tumor cell proliferation and survival, and angiogenesis[37]. Lenvatinib reduces tumor angiogenesis by inhibiting the vascular endothelial growth factor receptor and fibroblast growth factor receptor signaling pathways, indirectly downregulating the activity of the PI3K/Akt/mTOR signaling pathway[38]. Compared to those with a high TyG index, patients with a low TyG index have lower insulin resistance, and the PI3K/Akt/mTOR pathway is inherently less active. This synergistic inhibition could lead to apparent therapeutic effects and improved prognosis. A high TyG index was closely associated with obesity. Previous retrospective studies of colorectal cancer and malignant melanoma proposed that patients with obesity receiving immune checkpoint inhibitors had a better prognosis[39,40], a phenomenon known as the ‘obesity paradox’. In addition to HCC, the prognostic value of the TyG index has been explored in other malignancies. For instance, in colorectal cancer, a high TyG index has been associated with increased tumor invasiveness and poorer survival outcomes, likely due to its link with insulin resistance and systemic inflammation[12]. Similarly, studies in breast cancer populations have shown that elevated TyG levels correlate with worse prognosis, especially in patients with concurrent metabolic syndrome[41]. In contrast, findings in metastatic prostate cancer suggested that a lower TyG index was associated with improved response to immune checkpoint inhibitors, possibly due to reduced PD-1 expression and low immune suppression in these patients[15]. These findings indicate that while the TyG index is broadly relevant as a metabolic prognostic marker across cancers, its predictive strength and directionality may vary depending on tumor biology, immune microenvironment, and treatment context. Our study extends this knowledge by being the first to confirm the TyG index as an independent predictor of survival in patients with HBV-related HCC treated with combined immunotherapy and targeted therapy. Our results disagree with this phenomenon, possibly due to differences in tumor types, stages, and selected populations between our study and those in which it was described. Furthermore, the limitation of BMI lies in its inability to account for the effect of metabolic syndromes such as diabetes on immunotherapy outcomes. Therefore, the TyG index is likely to serve as a potential predictive biomarker.
Although the overall treatment response in patients with a high TyG index is lower than in those with a low TyG index, their risk of developing hypertension during treatment is higher. The higher incidence of hypertension observed in the high TyG index group might be partially explained by vascular tone dysregulation associated with insulin resistance and metabolic syndrome. Insulin resistance promotes endothelial dysfunction, reduces nitric oxide availability, and increases sympathetic activity, all of which contribute to elevated blood pressure[42]. Lenvatinib, a vascular endothelial growth factor receptor inhibitor, impairs endothelial repair and vasodilation, potentially exacerbating this effect in metabolically compromised patients[43]. Our findings suggest the need for proactive blood pressure monitoring and cardiovascular risk assessment, particularly in patients with a high TyG index receiving anti-angiogenic therapy. Clinicians should tailor management strategies based on the TyG index, paying particular attention to blood pressure changes in patients with a high TyG index and optimizing therapeutic approaches to minimize the risk of complications. However, the observed association should be interpreted with caution, as we did not perform multivariate adjustments for potential confounders such as age, baseline blood pressure, and comorbidities. Future studies with larger cohorts are needed to confirm whether the increased incidence of hypertension in the high TyG group remains significant after controlling for these variables.
We developed a prediction nomogram that encompasses the critical predictors of OS. Our analysis showed that the TyG index was the most influential predictor of OS. Other notable predictors included the BCLC stage, ECOG performance status, distant metastasis, and AFP level. Validation of the nomogram model showed a C-index of 0.81 in the training set and 0.78 in the internal validation set. Furthermore, calibration curves indicated a good agreement between the model’s predictions and the actual data, confirming the model’s reliability and accuracy. The nomogram developed in this study could be used clinically for individualized risk stratification. Since the TyG index is easily derived from routine laboratory tests, it can be incorporated into initial assessments to help identify patients at a higher risk of poor outcomes. Clinicians could consider closely monitoring metabolic comorbidities and blood pressure in patients with a high TyG index and explore interventions such as lifestyle modification or metabolic support to improve treatment response.
Our study had some limitations. First, this was a single-center retrospective study, which might be subject to selection bias and limited generalizability. Therefore, future validation should rely on multicenter prospective studies. Second, we chose the combination of lenvatinib and camrelizumab for our study to reduce treatment bias as much as possible. Although this selection strengthened the reliability of our analysis, it also implies that our findings need to be verified to ensure their generalizability to other PD-1/programmed death ligand-1 inhibitors and targeted therapies. Third, this study used a specific statistical method to determine the optimal cut-off value of the TyG index. However, this cut-off value could be influenced by factors such as sample size and the study population characteristics (e.g., ethnicity, age, and basal metabolic state), so it might not apply to other populations or medical contexts. Additionally, the TyG index calculation formulas and cut-off values might differ among studies, further limiting cross-study comparisons and the generalizability of the results. Therefore, larger-scale, multicenter studies are needed to validate the stability and universality of this cut-off value and explore whether there are stratified dynamic cut-off values to better guide clinical decisions. Fourth, 41 patients were excluded from the study because of missing pre-treatment TyG index data. We excluded these cases instead of performing imputation to ensure data integrity and avoid introducing bias from estimated values. However, this may have introduced a selection bias, as patients with missing data might differ systematically from those included. Moreover, we excluded patients with prior exposure to PD-1/programmed death ligand-1 inhibitors to reduce confounders arising from previous immunotherapy effects. While this improved internal validity, it may have limited the generalizability of our findings to immunotherapy-naïve populations. In real-world settings, many patients may have received prior lines of treatment. Therefore, future studies should include more diverse populations to evaluate whether the TyG index retains its prognostic value across treatment histories. Fifth, we did not explore potential interaction effects between the TyG index and other clinical variables, such as the BCLC stage or ECOG score. Future studies should consider including interaction terms in the multivariate models to assess whether the prognostic value of the TyG index varies across clinical subgroups. Finally, the single-center retrospective study characteristics and the sample size limitation hindered our ability to perform external validation.
Our prediction model could be integrated into the electronic health records as a pre-treatment risk stratification tool to facilitate clinical implementation, enabling clinicians to quickly identify patients who may require more intensive monitoring or tailored therapeutic approaches. The simplicity of the model, which is based on routinely available clinical variables such as the TyG index, makes it feasible for use in multidisciplinary treatment planning and real-time decision-making workflows. We acknowledge that external validation is essential to enhance the model’s generalizability. Therefore, we plan follow-up prospective multicenter studies with more diverse metabolic profiles to further validate and refine the nomogram.
CONCLUSION
Our study demonstrated that the TyG index is a significant prognostic biomarker for patients with HBV-related HCC undergoing combination therapy with lenvatinib and camrelizumab. A low TyG index was associated with improved long-term survival outcomes, suggesting its potential utility in guiding clinical decision-making and personalized treatment strategies for this patient population.
Footnotes
Provenance and peer review: Unsolicited article; Externally peer reviewed.
Peer-review model: Single blind
Specialty type: Gastroenterology and hepatology
Country of origin: China
Peer-review report’s classification
Scientific Quality: Grade A, Grade A, Grade A, Grade A, Grade B
Novelty: Grade A, Grade A, Grade A, Grade A, Grade B
Creativity or Innovation: Grade A, Grade A, Grade A, Grade B, Grade B
Scientific Significance: Grade A, Grade A, Grade A, Grade A, Grade B
P-Reviewer: Liu SC; Ma X; Tian G S-Editor: Wang JJ L-Editor: A P-Editor: Zheng XM
Wang Y, Lau W, Li Y, Tian Y, Lei Y, Wang J, Xia F. Clinico-characteristics of patients which correlated with preferable treatment outcomes in immunotherapy for advanced hepatocellular carcinoma: a systematic review and meta-analysis.Int J Surg. 2023;109:3590-3601.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 4][Reference Citation Analysis (0)]
Finn RS, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, Kudo M, Breder V, Merle P, Kaseb AO, Li D, Verret W, Xu DZ, Hernandez S, Liu J, Huang C, Mulla S, Wang Y, Lim HY, Zhu AX, Cheng AL; IMbrave150 Investigators. Atezolizumab plus Bevacizumab in Unresectable Hepatocellular Carcinoma.N Engl J Med. 2020;382:1894-1905.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 2542][Cited by in RCA: 4659][Article Influence: 931.8][Reference Citation Analysis (2)]
Wang X, Cao W, Qiu Y, Ji H, Yuan J, Wu W, Liu F, Feng L, Ding R, Li X, Tao K. Clinical efficacy and safety evaluation of camrelizumab plus lenvatinib in adjuvant therapy after hepatocellular carcinoma surgery.Front Oncol. 2023;13:1174999.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 2][Reference Citation Analysis (0)]
Fulgenzi CAM, Scheiner B, Korolewicz J, Stikas CV, Gennari A, Vincenzi B, Openshaw MR, Silletta M, Pinter M, Cortellini A, Scotti L, D'Alessio A, Pinato DJ. Efficacy and safety of frontline systemic therapy for advanced HCC: A network meta-analysis of landmark phase III trials.JHEP Rep. 2023;5:100702.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 20][Cited by in RCA: 34][Article Influence: 17.0][Reference Citation Analysis (0)]
Chang CH, Qiu J, O'Sullivan D, Buck MD, Noguchi T, Curtis JD, Chen Q, Gindin M, Gubin MM, van der Windt GJ, Tonc E, Schreiber RD, Pearce EJ, Pearce EL. Metabolic Competition in the Tumor Microenvironment Is a Driver of Cancer Progression.Cell. 2015;162:1229-1241.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1558][Cited by in RCA: 2278][Article Influence: 227.8][Reference Citation Analysis (0)]
Zhou J, Zhang M, Ju X, Wang H, Xiao H, Zhai Z, Zhong X, Hong J. Increased monocytic myeloid-derived suppressor cells in type 2 diabetes correlate with hyperglycemic and was a risk factor of infection and tumor occurrence.Sci Rep. 2024;14:4384.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 4][Reference Citation Analysis (0)]
Sueta D, Suyama K, Sueta A, Tabata N, Yamashita T, Tomiguchi M, Takeshita T, Yamamoto-Ibusuki M, Yamamoto E, Izumiya Y, Kaikita K, Yamamoto Y, Hokimoto S, Iwase H, Tsujita K. Lenvatinib, an oral multi-kinases inhibitor, -associated hypertension: Potential role of vascular endothelial dysfunction.Atherosclerosis. 2017;260:116-120.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 25][Cited by in RCA: 38][Article Influence: 4.8][Reference Citation Analysis (0)]
