Guo R, Gao M. Predictive value of neutrophil-lymphocyte ratio for prognosis in patients with advanced hepatocellular carcinoma treated with programmed cell death 1 inhibitors. World J Gastroenterol 2026; 32(18): 118416 [DOI: 10.3748/wjg.v32.i18.118416]
Corresponding Author of This Article
Rui Guo, Department of Laboratory Medicine, Henan Provincial People's Hospital, People's Hospital of Zhengzhou University, No. 7 Weiwu Road, Jinshui District, Zhengzhou 462000, Henan Province, China. guorui2460@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/
May 14, 2026 (publication date) through May 6, 2026
Times Cited of This Article
Times Cited (0)
Journal Information of This Article
Publication Name
World Journal of Gastroenterology
ISSN
1007-9327
Publisher of This Article
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Guo R, Gao M. Predictive value of neutrophil-lymphocyte ratio for prognosis in patients with advanced hepatocellular carcinoma treated with programmed cell death 1 inhibitors. World J Gastroenterol 2026; 32(18): 118416 [DOI: 10.3748/wjg.v32.i18.118416]
World J Gastroenterol. May 14, 2026; 32(18): 118416 Published online May 14, 2026. doi: 10.3748/wjg.v32.i18.118416
Predictive value of neutrophil-lymphocyte ratio for prognosis in patients with advanced hepatocellular carcinoma treated with programmed cell death 1 inhibitors
Rui Guo, Department of Laboratory Medicine, Henan Provincial People's Hospital, People's Hospital of Zhengzhou University, Zhengzhou 462000, Henan Province, China
Meng Gao, Department of Anesthesiology and Perioperative Medicine, Henan Cancer Hospital, Affiliated Cancer Hospital of Zhengzhou University, Zhengzhou 450000, Henan Province, China
Author contributions: Guo R initiated research; Meng Gao designed the experiments and conducted clinical data collection, performed postoperative follow-up and recorded the data; Guo R conducted the collation and statistical analysis, and wrote the original manuscript and revised the paper; all authors read and approved the final manuscript.
Institutional review board statement: This study was approved by the Ethics Committee of Henan Provincial People's Hospital, No. 2024-1-112.
Informed consent statement: The ethics committee agrees to waive informed consent.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: All data generated or analyzed during this study are included in this published article.
Corresponding author: Rui Guo, Department of Laboratory Medicine, Henan Provincial People's Hospital, People's Hospital of Zhengzhou University, No. 7 Weiwu Road, Jinshui District, Zhengzhou 462000, Henan Province, China. guorui2460@163.com
Received: January 9, 2026 Revised: January 30, 2026 Accepted: March 17, 2026 Published online: May 14, 2026 Processing time: 116 Days and 24 Hours
Abstract
BACKGROUND
The neutrophil-lymphocyte ratio (NLR) is an accessible inflammatory biomarker with emerging prognostic value in oncology.
AIM
To evaluate the predictive value of pretreatment NLR in patients with advanced hepatocellular carcinoma (aHCC) undergoing therapy with programmed cell death 1 (PD-1) inhibitors.
METHODS
This retrospective analysis included 234 patients with aHCC who received PD-1 inhibitor therapy between January 2022 and June 2024. The patients were categorized into good or poor prognosis groups according to overall survival relative to the median. The optimal NLR cutoff (3.165) was determined by receiver operator characteristic analysis, defining the low (≤ 3.165) and high (> 3.165) NLR groups. Hematological parameters were measured from blood samples collected within 1 week before treatment initiation. Radiologic assessments were conducted every 6-9 weeks in accordance with Response Evaluation Criteria in Solid Tumors v1.1. Progression-free survival (PFS) and overall survival (OS) were assessed during follow-up.
RESULTS
The high NLR group had significantly worse tumor burden and more advanced disease compared with the low NLR group. Multivariate analysis identified high NLR as an independent risk factor for poor prognosis (odds ratio = 4.365, P < 0.001) with the highest predictive accuracy (area under the curve = 0.785). The low NLR group demonstrated superior objective response rate (35.66% vs 16.19%, P = 0.001) and disease control rate (71.32% vs 57.14%, P = 0.024) and had significantly longer median progression-free survival (7.52 months vs 5.21 months, P < 0.001) and overall survival (16.84 months vs 11.05 months, P < 0.001). Cox regression confirmed that high NLR independently predicted poor PFS (hazard ratio = 2.084) and OS (hazard ratio = 2.421).
CONCLUSION
Pretreatment NLR is a powerful, independent prognostic biomarker for patients with aHCC receiving PD-1 inhibitor treatment.
Core Tip: This study demonstrates that a high pretreatment neutrophil-lymphocyte ratio (NLR) is an independent predictor of poor response and survival in advanced hepatocellular carcinoma patients treated with programmed cell death 1 inhibitors. NLR reflects a systemic inflammatory state that may undermine immunotherapy efficacy. Our findings support the use of this simple, cost-effective biomarker for risk stratification and treatment personalization in the immunotherapy era.
Citation: Guo R, Gao M. Predictive value of neutrophil-lymphocyte ratio for prognosis in patients with advanced hepatocellular carcinoma treated with programmed cell death 1 inhibitors. World J Gastroenterol 2026; 32(18): 118416
Hepatocellular carcinoma (HCC) represents a major global health challenge, ranking as a leading cause of cancer-related mortality worldwide[1,2]. Its high incidence, particularly in regions endemic for hepatitis B and C viruses, is compounded by frequent diagnosis at advanced stages when curative treatment options are no longer viable[3]. For decades, the therapeutic landscape for advanced HCC (aHCC) has remained limited, with multikinase inhibitors such as sorafenib offering only modest survival benefits[4,5]. This historical context underscores a significant unmet medical need that has long characterized the management of this aggressive malignancy.
The advent of immunotherapy has fundamentally transformed this treatment paradigm. Immune checkpoint inhibitors, particularly those targeting the programmed cell death protein 1 (PD-1) pathway, have emerged as a cornerstone of systemic therapy for aHCC[6,7]. By inhibiting PD-1 receptor on T-cells, these agents reinvigorate host’s antitumor immune response, leading to durable tumor control and improved survival for a subset of patients[8,9]. Drugs such as sintilimab, camrelizumab, and toripalimab have demonstrated substantial clinical efficacy and are now established in clinical practice, offering new hope for patients facing this formidable disease[10,11].
Despite these advances, a critical challenge persists, namely, the pronounced heterogeneity in treatment response. Some patients experience profound and long-lasting remission, whereas others derive minimal benefit, facing unnecessary toxicity and delays in transitioning to alternative therapies[12]. This stark dichotomy highlights the urgent need for reliable, easily accessible biomarkers capable of predicting outcomes prior to treatment initiation[13]. Such predictive tools are essential for optimizing patient selection, maximizing the cost-effectiveness of these powerful but expensive agents, and ultimately guiding personalized treatment strategies for aHCC.
In the search for practical prognostic indicators, the systemic inflammatory response has gained considerable attention as a key player in cancer pathogenesis and progression. The tumor microenvironment is heavily influenced by inflammatory cells, with neutrophils often exerting protumorigenic effects by promoting angiogenesis and suppressing immune activity and lymphocytes, particularly cytotoxic T-cells, being crucial for antitumor immunity[14,15]. The neutrophil-to-lymphocyte ratio (NLR), a simple, cost-effective, and readily available parameter derived from routine complete blood counts, quantitatively reflects this balance between protumor inflammation and antitumor immune competence[16,17].
The prognostic utility of NLR has been extensively documented across a wide spectrum of solid tumors, with elevated baseline levels consistently associated with poor survival outcomes[18]. In HCC, a disease frequently arising on a background of chronic liver inflammation, the interplay between inflammation and carcinogenesis is particularly pronounced. Although preliminary investigations indicated a possible association between high NLR and adverse prognosis in patients with HCC receiving various treatments, its specific role and predictive power in the contemporary setting of PD-1 inhibitor therapy remain to be fully elucidated and validated. Therefore, this study aims to comprehensively investigate the predictive value of pretreatment NLR for prognosis in patients with aHCC treated with PD-1 inhibitors.
MATERIALS AND METHODS
Study population
In this retrospective study, 234 patients with HCC who underwent PD-1 inhibitor therapy at Henan Provincial People’s Hospital between January 1, 2022, and June 30, 2024 were enrolled. Adult patients (age ≥ 18 years) with a confirmed diagnosis of aHCC[19] and who had undergone a minimum of two cycles of first-line PD-1 inhibitor treatment were included. Eligible participants were required to have complete blood routine test data available within 7 days prior to treatment initiation, an Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 0-1, and a Child-Pugh A or B grade liver function score of ≤ 7. Patients were excluded from the analysis if they had concurrent other active malignancies, coexisting active infections or autoimmune diseases, or were on systemic immunosuppressants. Individuals lacking complete clinical records or insufficient follow-up data were also excluded.
The study protocol received approval from the Institutional Review Board of Henan Provincial People’s Hospital. Owing to the retrospective nature of the research, the requirement for obtaining informed consent was waived. All patient records were processed with strict confidentiality to ensure the protection of personally identifiable information throughout the study.
Patient grouping
The participants were categorized into a good prognosis group (n = 126) and a poor prognosis group (n = 108) according to whether their overall survival (OS) was longer than the median survival time (12 months). The optimal threshold for NLR was established through receiver operating characteristic (ROC) curve analysis and identified as 3.165. On the basis of this threshold, the entire cohort was further subdivided into a low NLR group (≤ 3.165, n = 129) and a high NLR group (> 3.165, n = 105).
Therapeutic regimen and efficacy assessment
All patients received PD-1 inhibitor monotherapy as first-line systemic therapy. No patients in this cohort received concurrent anti-angiogenic agents or other systemic therapies in combination with PD-1 inhibitors during the study period. The specific agents used included sintilimab [200 mg intravenously (iv) every 3 weeks], camrelizumab (200 mg iv every 3 weeks), or toripalimab (240 mg iv every 3 weeks) in accordance with Chinese national guidelines and institutional standards. Treatment cycles were repeated every 3 weeks, and patients were required to have received at least two cycles to be eligible for inclusion. Treatment was administered until disease progression was confirmed radiologically, unacceptable toxicity or death occurred, or the patient decided to withdraw consent.
Tumor response was evaluated using contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI), and the outcomes were analyzed based on the Response Evaluation Criteria in Solid Tumors version 1.1. Complete response (CR) was characterized by the disappearance of all target lesions. Partial response (PR) was identified as at least a 30% reduction in the total diameter of target lesions. Stable disease (SD) was characterized by neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for progressive disease (PD). PD was defined as at least a 20% increase in the sum of diameters of target lesions or the appearance of new lesions. Objective response rate (ORR) was calculated as the proportion of patients achieving CR or PR. Disease control rate (DCR) was calculated as the proportion of patients achieving CR, PR, or SD[20].
Data collection
Clinical, demographic, and laboratory data were retrospectively collected from the hospital’s electronic medical record system.
Baseline characteristics: Baseline characteristics recorded at the time of treatment initiation included gender, age, body mass index (BMI), past tobacco use and alcohol intake, etiology of liver disease (categorized as hepatitis B, hepatitis C, or others), Barcelona Clinic Liver Cancer (BCLC) stage, Child-Pugh grade, ECOG PS score, maximum tumor diameter (measured on pretreatment imaging), presence of macrovascular invasion, and presence of extrahepatic spread.
Hematological parameters: Peripheral venous blood samples were obtained from all participants within 7 days prior to the start of PD-1 inhibitor therapy. Neutrophil, lymphocyte, and platelet counts were analyzed using an automated hematology analyzer (Sysmex XN-9000, Sysmex Corporation, Japan). NLR was calculated by dividing the absolute neutrophil count by the absolute lymphocyte count. Parameters including albumin, total bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen, and serum creatinine were measured using a fully automated biochemical analyzer (Cobas c 702, Roche Diagnostics, Switzerland). International normalized ratio (INR) was determined using a coagulation analyzer (CS-5100, Sysmex Corporation, Japan). Serum alpha-fetoprotein (AFP) level was quantified using an electrochemical luminescence immunoassay system (Cobas e 602, Roche Diagnostics, Switzerland).
Follow-up
Follow-up with patients was performed via outpatient clinic visits and telephone interviews. The follow-up protocol began on the date of the initial PD-1 inhibitor infusion. Radiological tumor assessments (via CT or MRI) were performed every 6-9 weeks (approximately every 2-3 treatment cycles) to evaluate treatment response. The median follow-up time for the entire cohort was 14.2 months (range: 6.0-34.8 months). At the final follow-up date (December 31, 2024), 147 patients (62.8%) were still alive, and 187 patients (79.9%) had experienced disease progression or death. Progression-free survival (PFS) was characterized as the duration from the start of PD-1 inhibitor treatment to the initial confirmed radiological progression or death due to any reason. OS was characterized as the duration from the start of treatment to death from any cause.
Statistical analysis
Data analyses were conducted with SPSS 29.0 (IBM Corp, United States). Shapiro-Wilk test confirmed that all continuous variables followed a normal distribution. Continuous variables were presented as mean ± SD and compared using independent samples t-test. Categorical variables were expressed as n (%) and compared using χ2 test. Multivariate logistic regression analysis was employed to identify independent risk factors associated with poor prognosis, with results presented as odds ratio (OR) and 95% confidence interval (CI). The predictive performance of various parameters for poor prognosis was assessed using the ROC curve, and the area under the curve (AUC), sensitivity, specificity, Youden index, and F1 score were calculated. Survival analysis for PFS and OS between the low and high NLR groups was conducted using the Kaplan-Meier method and compared using log-rank test. Univariate and multivariate Cox proportional hazards regression models were used to determine factors independently linked with PFS and OS, with results reported as hazard ratio (HR) and 95%CI. All tests were conducted on a two-sided basis. P < 0.05 was considered statistically significant.
RESULTS
Baseline information of patients with different prognostic statuses
By comparing their baseline characteristics, we found no significant differences in gender distribution (P = 0.851), age (P = 0.309), BMI (P = 0.126), history of smoking (P = 0.455), history of alcohol consumption (P = 0.528), etiology of liver disease (P = 0.883), and Child-Pugh grade (P = 0.083) between the patients with aHCC showing good and poor prognosis treated with PD-1 inhibitors. However, the BCLC stage of the poor prognosis group was markedly greater than that of the good prognosis group (χ2 = 5.661, P = 0.017). The ECOG PS score of the good prognosis group showed a higher reduction (χ2 = 3.951, P = 0.047). The maximum tumor diameter of the poor prognosis group was larger (t = 4.350, P < 0.001). The macrovascular invasion rates were greater (χ2 = 6.588, P = 0.010) and the extrahepatic spread was also more frequent (χ2 = 5.091, P = 0.024) in the poor prognosis group (Table 1). These findings indicate that patients with poor prognosis have more advanced stages of cancer, larger tumors, higher prevalence of macrovascular invasion, and more frequent extrahepatic metastasis compared with those having a good prognosis.
Table 1 Comparison of baseline information between patients with good prognosis and poor prognosis, n (%)/mean ± SD.
Hematological parameters of patients with different prognostic statuses
In Table 2, we present a comparison of hematological parameters between patients with good and poor prognosis treated with PD-1 inhibitors for aHCC. The AFP levels of the good prognosis group were markedly lower than that of the poor prognosis group (t = 8.248, P < 0.001). NLR was also significantly lower in the good prognosis group (t = 8.296, P < 0.001). The poor prognosis group exhibited greater levels of AST (t = 2.438, P = 0.016) and total bilirubin (t = 2.943, P = 0.004). Albumin levels were markedly higher in the good prognosis group (t = 3.046, P = 0.003), and INR was higher in the poor prognosis group (t = 2.928, P = 0.004). No marked distinctions in ALT (P = 0.105), blood urea nitrogen (P = 0.180), or serum creatinine (P = 0.167) were observed between the two groups (Table 2). These results indicate that the patients with good prognosis have lower levels of AFP, NLR, AST, total bilirubin, and INR but higher albumin levels compared with those having a poor prognosis.
Table 2 Comparison of hematological parameters between patients with good prognosis and poor prognosis, mean ± SD.
Multivariate logistic regression analysis identified several factors influencing poor prognosis (Table 3). Higher maximum tumor diameter (P = 0.004, OR = 1.331) and AFP (P < 0.001, OR = 1.006) were markedly correlated with a heightened probability of poor prognosis. Elevated NLR was also a strong predictor of poor prognosis (P < 0.001, OR = 4.365). High AST levels (P = 0.012, OR = 1.039), and total bilirubin levels (P = 0.010, OR = 1.119) were linked with poor outcomes. Conversely, high albumin levels showed a protective effect against poor prognosis (P = 0.012, OR = 0.880). Compared with B, BCLC stage C showed a significant association with poor prognosis (P = 0.048, OR = 2.336), indicating that advanced cancer stages are predictive of poor outcomes. Factors such as macrovascular invasion (P = 0.248), extrahepatic spread (P = 0.150), and INR (P = 0.545) did not show statistically significant associations with prognosis in this multivariate model (Table 3).
ROC analysis evaluated the diagnostic performance of various factors influencing poor prognosis in patients with aHCC treated with PD-1 inhibitors (Table 4). Among these factors, NLR demonstrated the highest AUC value (AUC = 0.785, Best_threshold = 3.165), indicating its strong predictive capability for poor prognosis. AFP levels also showed a relatively high AUC (AUC = 0.778), suggesting that it is another valuable biomarker for prognosis prediction. Maximum tumor diameter had an AUC of 0.657, indicating moderate predictive power. AST levels showed lower discriminative ability (AUC = 0.586). Total bilirubin and albumin levels exhibited similar AUC values (total bilirubin: AUC = 0.598; albumin: AUC = 0.607), which were slightly better than those of AST but still moderate. BCLC stage had the lowest AUC among all parameters evaluated (AUC = 0.567), indicating its limited predictive utility for distinguishing between good and poor prognosis. These findings reinforce NLR as a key independent prognostic factor in this patient population (Table 4 and Figure 1).
Baseline characteristics of patients in different NLR groups
By comparing baseline characteristics between the low and high NLR groups, we observed several significant differences. The two groups showed no marked distinctions in gender distribution (P = 0.474), age (P = 0.315), BMI (P = 0.224), history of smoking (P = 0.259), history of alcohol consumption (P = 0.329), or etiology of liver disease (P = 0.906). However, marked distinctions were noted for BCLC stage (χ2 = 5.828, P = 0.016), Child-Pugh grade approached significance (P = 0.052), and ECOG PS score (χ2 = 4.081, P = 0.043). The maximum tumor diameter was significantly larger in the high NLR group (t = 4.800, P < 0.001). Rates of macrovascular invasion (χ2 = 10.593, P = 0.001) and extrahepatic spread (χ2 = 6.673, P = 0.010) were also higher in the high NLR group than in the low NLR group. These findings indicate that patients with high NLR have advanced stages of cancer and large tumors and are likely to experience macrovascular invasion and extrahepatic metastasis (Table 5).
Table 5 Comparison of baseline characteristics between patients with low neutrophil-to-lymphocyte ratio and high neutrophil-to-lymphocyte ratio, n (%)/mean ± SD.
Hematological parameters of patients in different NLR groups
In Table 6, we compared hematological parameters between the low and high NLR groups and observed several significant differences. AFP levels were significantly lower in the low NLR group than in the high NLR group (t = 10.26, P < 0.001). AST (t = 2.812, P = 0.005) and total bilirubin (t = 2.983, P = 0.003) levels were higher in the high NLR group. Albumin levels were significantly higher in the low NLR group (t = 3.099, P = 0.002). INR was higher in the high NLR group (t = 3.472, P < 0.001). No significant differences in ALT (P = 0.131), blood urea nitrogen (P = 0.093), or serum creatinine (P = 0.176) were observed between the two groups. These results indicate that patients with high NLR have elevated AFP, AST, total bilirubin, and INR levels but lower albumin levels compared with those having low NLR (Table 6).
Table 6 Comparison of hematological parameters between patients with low neutrophil-to-lymphocyte ratio and high neutrophil-to-lymphocyte ratio.
Clinical efficacy of patients in different NLR groups
Marked distinctions in clinical outcomes were observed between the two groups. Compared with the high NLR group, the low NLR group exhibited greater ORR, comprising CR and PR (χ2 = 10.386, P = 0.001). The DCR, encompassing CR, PR, and SD, was also markedly greater in the low NLR group (χ2 = 5.110, P = 0.024). These data indicate that patients with low NLR have good responses to PD-1 inhibitor treatment, as evidenced by their high ORR and DCR (Table 7).
Table 7 Comparison of efficacy between patients with low neutrophil-to-lymphocyte ratio and high neutrophil-to-lymphocyte ratio, n (%).
To further evaluate whether NLR independently predicts treatment response beyond tumor burden and stage, we performed a multivariate logistic regression analysis with ORR (CR + PR) as the dependent variable. After adjusting for BCLC stage (C vs B), maximum tumor diameter, macrovascular invasion, and extrahepatic spread, we found that high NLR remained a significant independent predictor of low ORR (adjusted OR = 0.352, 95%CI: 0.182-0.680, P = 0.002). In this model, other tumor burden variables were not significantly associated with ORR. This finding indicates that the association between NLR and treatment response is not merely reflective of disease severity, but rather an independent biomarker of immunotherapy efficacy (Table 8).
In Figure 2, we compared PFS and OS between patients with low and high NLR treated with PD-1 inhibitors for aHCC. The results demonstrate significant differences in PFS and OS between the two groups. Low NLR patients had a markedly longer PFS than high NLR patients (t = 8.434, P < 0.001). Similarly, the low NLR group exhibited a significantly longer OS (t = 11.052, P < 0.001). These findings indicate that patients with low NLR have good survival outcomes in terms of PFS and OS when undergoing therapy with PD-1 inhibitors. Therefore, NLR may serve as indicator for estimating survival in patients with aHCC (Figure 2).
Figure 2 Kaplan-Meier plots of progression-free survival and overall survival between patients with low neutrophil-to-lymphocyte ratio and high neutrophil-to-lymphocyte ratio.
A: Progression-free survival (PFS) curve; B: PFS patients number at risk; C: Overall survival (OS) curve; D: OS patients number at risk. NLR: Neutrophil-to-lymphocyte ratio.
Cox proportional hazards regression analysis of the impact on survival outcome
In our Cox proportional hazards regression analysis examining factors influencing PFS in patients with aHCC treated with PD-1 inhibitors, several significant predictors were identified. In the univariate analysis, BCLC stage C was associated with a significantly higher risk of disease progression (HR = 1.824, P = 0.002) compared with B, and this risk remained markedly high in the multivariate model (HR = 1.512, P = 0.040). Maximum tumor diameter per 1 cm increase also showed a significant association with poor PFS in univariate (HR = 1.283, P < 0.001) and multivariate (HR = 1.191, P = 0.001) analyses. AFP levels ≥ 400 ng/mL were another significant predictor of poor PFS in univariate (HR = 1.962, P < 0.001) and multivariate (HR = 1.623, P = 0.004) models. High NLR emerged as one of the strongest predictors of poor PFS, with a significant association in univariate (HR = 2.451, P < 0.001) and multivariate (HR = 2.084, P < 0.001) analyses. Other parameters such as Child-Pugh grade B vs A (univariate HR = 1.521, P = 0.014; multivariate P = 0.165), macrovascular invasion (univariate HR = 1.654, P = 0.002; multivariate P = 0.096), extrahepatic spread (univariate HR = 1.584, P = 0.006; multivariate P = 0.188), AST levels ≥ 40 U/L (univariate HR = 1.421, P = 0.028; multivariate P = 0.184), total bilirubin levels ≥ 20 μmol/L (univariate HR = 1.378, P = 0.045; multivariate P = 0.325), and albumin levels ≥ 35 g/L (univariate HR = 0.721, P = 0.042; multivariate P = 0.216) did not maintain their significance in the multivariate model. The strong association between high NLR and poor PFS suggests the potential utility of NLR as a prognostic biomarker (Table 9).
Table 9 Cox proportional hazards regression analysis of factors influencing progression-free survival in patients with advanced hepatocellular carcinoma treated with programmed cell death protein 1 inhibitors.
In the Cox proportional hazards regression analysis examining factors influencing OS in patients with aHCC undergoing therapy with PD-1 inhibitors, several significant predictors were identified. In the univariate analysis, BCLC stage C showed a markedly higher risk of mortality (HR = 2.148, P < 0.001) compared with B, and this association remained significant in the multivariate model (HR = 1.763, P = 0.010). Maximum tumor diameter per 1 cm increase also demonstrated a significant association with poor OS in univariate (HR = 1.352, P < 0.001) and multivariate (HR = 1.241, P < 0.001) analyses. AFP levels ≥ 400 ng/mL were another significant predictor of poor OS in univariate (HR = 2.241, P < 0.001) and multivariate (HR = 1.847, P = 0.001) models. High NLR emerged as one of the strongest predictors of poor OS, with a significant association in univariate (HR = 2.861, P < 0.001) and multivariate (HR = 2.421, P < 0.001) analyses. Other parameters such as Child-Pugh grade B vs A showed significance in the univariate analysis (HR = 1.681, P = 0.005) but did not reach statistical significance in the multivariate model (HR = 1.405, P = 0.077). Macrovascular invasion and extrahepatic spread were both significantly associated with poor OS in univariate (macrovascular invasion: HR = 1.892, P < 0.001; extrahepatic spread: HR = 1.821, P = 0.001) and multivariate (macrovascular invasion: HR = 1.475, P = 0.032; extrahepatic spread: HR = 1.521, P = 0.029) analyses. AST levels ≥ 40 U/L, total bilirubin levels ≥ 20 μmol/L, and albumin levels ≥ 35 g/L were significant in the univariate analysis, (AST: HR = 1.578, P = 0.009; total bilirubin: HR = 1.518, P = 0.016; albumin: HR = 0.652, P = 0.016) but lost significance in the multivariate model (P = 0.128, P = 0.202, P = 0.108, respectively). The strong association between high NLR and poor OS suggests the potential utility of NLR as a prognostic biomarker (Table 10).
Table 10 Cox proportional hazards regression analysis of factors influencing overall survival in patients with advanced hepatocellular carcinoma treated with programmed cell death protein 1 inhibitors.
aHCC management has been transformed by immune checkpoint inhibitors, yet the reliable prediction of treatment outcomes remains an elusive goal. Our investigation provides compelling evidence that pretreatment NLR serves as a powerful and independent prognostic biomarker in this patient population. The findings delineate a clear association between an elevated NLR and an aggressive disease phenotype, diminished therapeutic response, and inferior survival, underscoring its clinical utility in risk stratification.
Our initial analysis revealed that patients having poor prognosis presented with an advanced disease state, characterized by a high prevalence of advanced BCLC stage, impaired performance status, large tumor dimensions, and increased incidence of macrovascular invasion and extrahepatic spread[21,22]. This clinical profile is consistent with a tumor biology that is intrinsically aggressive and prone to dissemination. The correlation between these established markers of tumor burden and a poor outcome aligns with the extensive literature on HCC natural history, confirming that our cohort reflects the expected spectrum of disease severity[23,24].
Examination of hematological parameters further illuminated the distinct profiles between the prognostic groups. Patients with poor outcomes exhibited elevated levels of AFP, a well-established tumor marker reflective of biological aggressiveness in HCC[25]. They also demonstrated high NLR and alterations in liver function tests indicative of great hepatic compromise, such as increased AST and total bilirubin and decreased albumin[26]. The convergence of elevated AFP, a tumor-derived factor, with an elevated NLR, a host-derived inflammatory marker, suggests a synergistic interaction between aggressive tumor biology and a permissive systemic environment[27]. This combination appears to foster a milieu less conducive to effective immunotherapy[28].
The central finding of this study is the identification of NLR as an independent predictor of poor prognosis through multivariate analysis. Even after the influence of other powerful prognostic factors such as tumor size, AFP, and liver function parameters was considered, high NLR retained a strong and independent association with unfavorable outcomes[29]. This finding underscores that the prognostic information carried by NLR is not merely a surrogate for other known risk factors but represents a distinct biological pathway influencing patient survival. The predictive power of NLR was further affirmed by ROC analysis, which demonstrated its superior discriminatory capacity to other clinical and laboratory parameters, including the established BCLC staging system[18].
The biological plausibility of our findings is rooted in the evolving understanding of cancer-related inflammation. An elevated NLR quantitatively represents a systemic imbalance, favoring protumor neutrophilic inflammation over antitumor lymphocyte-mediated immunity[29]. Neutrophils promote tumor progression through various mechanisms, including the release of angiogenic factors, matrix-degrading enzymes, and reactive oxygen species that foster genetic instability[30]. Furthermore, they can directly suppress the cytotoxic activity of T-cells and natural killer cells, effectively blunting the immune response that PD-1 inhibitors seek to unleash. Conversely, lymphocytes, particularly cytotoxic T-cells, are the primary effectors of anti-tumor immunity[31]. A relative lymphocytopenia, as reflected by a high NLR, indicates an impaired immune surveillance apparatus[32]. Therefore, a high NLR likely identifies a patient subset with a profoundly immunosuppressive tumor microenvironment, where the preexisting barriers to immune activation are too formidable for PD-1 blockade to overcome effectively. In our study population predominantly composed of hepatitis B virus-related HCC, a high NLR might reflect a state of chronic, virus-driven inflammation that is systemic and pervasive within the liver. This state could be associated with a functionally exhausted immune landscape, potentially explaining the reduced efficacy of PD-1 inhibitors aimed at reinvigorating T-cell function.
Regarding future directions, NLR, as a systemic and readily available marker, may hold a complementary rather than substitutive relationship with tissue-based markers such as PD-L1. While PD-L1 reflects a specific checkpoint interaction at the tumor site, NLR provides a broad, host-level assessment of the inflammatory and immune context. Their combined use could offer a holistic view of the local and systemic antitumor immune status, warranting investigation in future studies. This mechanistic framework explains the observed resistance to immunotherapy in this group[33,34].
The clinical implications of this biological state are directly reflected in our efficacy and survival analyses. Patients with low NLR achieved objectively superior response rates and DCRs following PD-1 inhibitor therapy. Multivariate logistic regression confirmed that a high NLR independently predicted low odds of achieving an objective response even after adjustment for BCLC stage and tumor burden, underscoring that its predictive value for efficacy extends beyond being a mere surrogate for disease severity. This finding indicates that a favorable immune contexture, as denoted by a low NLR, is associated with a high likelihood of the tumor responding to immune checkpoint blockade[35]. Ultimately, this phenomenon is translated into a substantial survival benefit, with the low NLR group experiencing long PFS and OS. The Cox regression models solidified this observation, confirming that a high NLR independently predicts disease progression and mortality. The magnitude of this association suggests that NLR is a biomarker of considerable clinical importance[36]. Our results are consistent with a growing body of literature across various cancers treated with immunotherapy, where a high baseline NLR consistently portends poor outcomes[37,38]. Although previous research on HCC hinted at this relationship, our study provides a comprehensive analysis within a homogeneous cohort treated exclusively with contemporary PD-1 inhibitors, strengthening the evidence for its specific application in the modern immunotherapy era.
Despite the clear associations demonstrated in this study, several limitations should be considered. The retrospective design and single-center context increase the risk of selection bias and reduce the external validity of the results. Our study focused on PD-1 inhibitor monotherapy. In contemporary practice, combination regimens (e.g., with anti-angiogenic agents) are increasingly used, which may influence baseline inflammatory profiles and treatment outcomes. Future studies should evaluate NLR in the context of combination therapies. The cohort was predominantly driven by hepatitis B etiology, and the applicability of our specific NLR cutoff value (3.165) to populations with different predominant etiologies, such as hepatitis C or nonalcoholic steatohepatitis, remains unknown. Different underlying inflammatory backgrounds (e.g., viral vs metabolic) could significantly influence baseline NLR distributions and its relationship with outcomes. Therefore, validation of this cutoff in external, multicenter cohorts, particularly those with diverse etiologies, is a critical step prior to its adoption as a clinically practical tool. The definition of prognostic groups based on median survival, while practical, is a methodological approach that should be corroborated using predefined, fixed-time survival endpoints in future research. Furthermore, our study focused on a single time point - baseline NLR. The dynamics of NLR during treatment and its potential utility as an on-treatment biomarker represent an intriguing area for future investigation.
Future research should be directed toward large-scale, prospective, multicenter studies to validate the prognostic cutoff value of NLR established in this work. Integrating NLR with other emerging biomarkers, such as PD-L1 expression, tumor mutational burden, or circulating tumor DNA, could lead to the development of a composite prognostic model with enhanced predictive power. Exploring the biological drivers behind a high NLR in nonresponders could also reveal novel therapeutic targets. For instance, combinations of PD-1 inhibitors with agents designed to modulate the tumor microenvironment by targeting neutrophil recruitment or function might represent a viable strategy to overcome resistance in this high-risk patient subset identified by an elevated NLR.
CONCLUSION
Our study robustly demonstrates that pretreatment NLR is a readily accessible and independent prognostic biomarker for patients with aHCC undergoing PD-1 inhibitor therapy. It reflects a deleterious balance between protumor inflammation and antitumor immunity, which in turn influences therapeutic response and survival outcomes. The integration of this simple, cost-effective parameter into treatment planning could contribute to optimizing patient selection and tailoring medical approaches in the era of immunotherapy.
Zhang L, Feng J, Kuang T, Chai D, Qiu Z, Deng W, Dong K, Zhao K, Wang W. Blood biomarkers predict outcomes in patients with hepatocellular carcinoma treated with immune checkpoint Inhibitors: A pooled analysis of 44 retrospective sudies.Int Immunopharmacol. 2023;118:110019.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 43][Reference Citation Analysis (0)]
Zhang S, Chen Z, Ling J, Feng Y, Xie Y, Liu X, Hu C, Hou T. Nomograms based on the lymphocyte-albumin-neutrophil ratio (LANR) for predicting the prognosis of nasopharyngeal carcinoma patients after definitive radiotherapy.Sci Rep. 2024;14:5388.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 5][Reference Citation Analysis (0)]
Iavarone M, Alimenti E, Canova L, Bruccoleri M, Antonelli B, Ierardi AM, Sangiovanni A, Cabibbo G, De Silvestri A, Caccamo L, Carrafiello G, Lampertico P. The impact of BCLC recommendations on survival for patients with hepatocellular carcinoma.Hepatol Commun. 2025;9:e0750.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 5][Reference Citation Analysis (0)]
Kim JH, Kim JH, Yoon HK, Ko GY, Shin JH, Gwon DI, Ko HK, Chu HH, Kim SH, Kim GH, Kim Y, Aljerdah S. Transarterial chemoembolization for advanced hepatocellular carcinoma without macrovascular invasion or extrahepatic metastasis: analysis of factors prognostic of clinical outcomes.Front Oncol. 2023;13:1072922.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in RCA: 4][Reference Citation Analysis (0)]
Zhu HF, Feng JK, Xiang YJ, Wang K, Zhou LP, Liu ZH, Cheng YQ, Shi J, Guo WX, Cheng SQ. Combination of alpha-fetoprotein and neutrophil-to-lymphocyte ratio to predict treatment response and survival outcomes of patients with unresectable hepatocellular carcinoma treated with immune checkpoint inhibitors.BMC Cancer. 2023;23:547.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 15][Reference Citation Analysis (0)]
Liu W, Shi T, Lu C, Che K, Zhang Z, Luo Y, Hirschhorn D, Wang H, Liu S, Wang Y, Liu S, Sun H, Lu J, Liu Y, Shi D, Ding S, Xu H, Lu L, Xu J, Xin J, Liang Y, Merghoub T, Wei J, Li Y. Human myelocyte and metamyelocyte-stage neutrophils suppress tumor immunity and promote cancer progression.Cell Res. 2025;35:588-606.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in RCA: 3][Reference Citation Analysis (0)]
Suzuki T, Matsuura K, Sobue S, Kato D, Hayashi K, Kondo H, Anbe K, Mizoshita T, Okayama K, Okumura F, Kimura Y, Ozasa A, Inada H, Tokunaga T, Narahara S, Kawamura H, Fujiwara K, Nojiri S, Kataoka H, Tanaka Y. Neutrophil-to-lymphocyte ratio at the start of the second course of durvalumab plus tremelimumab therapy predicts therapeutic efficacy in patients with advanced hepatocellular carcinoma: A multicenter analysis.Hepatol Res. 2025;55:883-895.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 4][Reference Citation Analysis (1)]
Tada T, Kumada T, Hiraoka A, Hirooka M, Kariyama K, Tani J, Atsukawa M, Takaguchi K, Itobayashi E, Fukunishi S, Tsuji K, Ishikawa T, Tajiri K, Ochi H, Yasuda S, Toyoda H, Ogawa C, Nishimura T, Hatanaka T, Kakizaki S, Shimada N, Kawata K, Tanaka T, Ohama H, Nouso K, Morishita A, Tsutsui A, Nagano T, Itokawa N, Okubo T, Arai T, Imai M, Naganuma A, Koizumi Y, Nakamura S, Joko K, Iijima H, Hiasa Y; Real-life Practice Experts for HCC (RELPEC) Study Group and the Hepatocellular Carcinoma Experts from 48 clinics in Japan (HCC 48) Group. Neutrophil-lymphocyte ratio predicts early outcomes in patients with unresectable hepatocellular carcinoma treated with atezolizumab plus bevacizumab: a multicenter analysis.Eur J Gastroenterol Hepatol. 2022;34:698-706.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 43][Cited by in RCA: 51][Article Influence: 12.8][Reference Citation Analysis (0)]
