Prognostic Factors Analysis in Patients with Hepatocellular Carcinoma Receiving PD-1/PD-L1 Inhibitors

Abstract

Background

Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of advanced hepatocellular carcinoma (HCC); however, only a minority of patients derive benefits, and reliable biomarkers remain lacking. This study aimed to investigate the associations between peripheral blood markers, clinical characteristics, and outcomes in patients receiving programmed cell death protein-1 (PD-1)/programmed death-ligand 1 (PD-L1) inhibitors.

Methods

Clinical and laboratory data from 101 patients treated with PD‑1/PD‑L1 inhibitors were retrospectively analyzed. Optimal cutoff values for continuous variables were determined using the X‑tile. Intergroup comparisons were performed using Pearson’s chi-squared or Fisher’s exact tests. Univariate Cox regression (SPSS 26.0) was used to identify potential prognostic factors (P<0.1), which were further evaluated by Log rank test and Kaplan–Meier curves. Independent prognostic factors were determined using a multivariate Cox regression analysis.

Results

The factors associated with efficacy included body mass index, Child‑Pugh class, line of treatment, distant metastasis, bone metastasis, alkaline phosphatase, and total bilirubin. Prognosis‑related indicators included lung metastasis, hemoglobin, absolute neutrophil count, alkaline phosphatase, gamma‑glutamyl transferase, C‑reactive protein, and the best response.

Conclusion

Based on routine clinical and hematological data, this prognostic study provides an accessible tool for personalized immunotherapy management of HCC.

Introduction

Hepatocellular carcinoma (HCC) is the most common primary malignancy of the liver, with a low 5-year relative survival rate, and is the third leading cause of cancer-related death worldwide.¹ Globally, there are over 900,000 new cases of liver cancer and more than 800,000 liver cancer-related deaths each year.² The incidence and mortality rates of liver cancer are gradually increasing in North America and some regions of Europe, while showing a decreasing trend in certain areas of China.³ Nevertheless, China still has more than 400,000 new cases of liver cancer annually, with over 380,000 liver cancer-related deaths each year.⁴ Compared to women, men have a higher risk of developing HCC.³ Viral hepatitis, alcohol consumption, and non-alcoholic fatty liver disease are the main risk factors for HCC. In addition, aflatoxin exposure, smoking, diabetes, and obesity are also associated with an increased incidence of HCC.⁵ Alpha-fetoprotein (AFP) is currently the most widely used serum biomarker for the diagnosis and evaluation of HCC.⁶ However, most patients with HCC are diagnosed at an advanced stage or with unresectable disease.⁷ Therefore, early diagnosis and treatment represent an effective strategy to improve patient prognosis.

Current treatment modalities for HCC mainly include surgical resection, radiotherapy, liver transplantation, transarterial chemoembolization (TACE), targeted therapy, immunotherapy, and combination regimens involving multiple antitumor agents.⁸ Hepatic resection and liver transplantation are the preferred options for curative treatment in patients with tumors; however, the high postoperative recurrence rate renders other therapeutic approaches indispensable.^8,9 Recent years have witnessed the emergence of novel treatment modalities such as TACE and radiofrequency ablation, which can effectively treat localized lesions but fail to completely eliminate residual cancer cells within the body, thereby predisposing patients to tumor recurrence and metastasis. Therefore, a combination therapy with other therapeutic approaches is warranted. Targeted therapies play an important role in treating advanced HCC. Agents such as sorafenib, lenvatinib, and ramucirumab have been used in the treatment of HCC.^10–13 Due to the genetic heterogeneity of HCC, these agents provide only limited benefits for patients.¹⁴

Immunotherapy can stimulate specific immune responses in the host, inhibit and kill tumor cells, and enhance the antitumor immune response of the body. Immunotherapy includes immune checkpoint inhibitors (ICIs), immune cell therapy, cytokine therapy, and tumor vaccines. Within the liver, various immune cells from both the innate and adaptive immune systems are present to clear endotoxins and invading pathogens.¹⁵ The advent of immune checkpoint inhibitors has broadened the therapeutic landscape for HCC. The IMbrave-150 study demonstrated that atezolizumab plus bevacizumab significantly improved overall survival (OS) and progression-free survival (PFS) compared with sorafenib in patients with unresectable HCC.¹⁶ This regimen has also been approved by the US Food and Drug Administration (FDA) for the treatment of unresectable locally advanced or metastatic HCC.¹⁷ The results of the CheckMate-040 study showed that nivolumab combined with ipilimumab achieved a median overall survival (mOS) of 22.8 months and an objective response rate (ORR) of 32% in patients with advanced HCC, indicating a significant therapeutic effect of this regimen for advanced HCC.¹⁸ A Phase II clinical trial conducted in China showed that camrelizumab achieved an ORR of 14.7% in patients with advanced hepatocellular carcinoma, with an overall survival rate of 74.4% at 6 months and 55.9% at 12 months. Accordingly, camrelizumab has been approved in China for the treatment of patients with advanced HCC who have previously received sorafenib and/or oxaliplatin-containing systemic chemotherapy.¹⁹

Primary or secondary resistance can render ICIs ineffective, leading to tumor relapse or metastasis. Tumor cell-intrinsic features, such as the interferon signaling pathway, the expression of antigen-presenting molecules, and immune evasion oncogenic signaling pathways, can affect T cell activation and inhibit their recruitment to the tumor microenvironment (TME), thereby causing resistance to ICIs in patients.²⁰ Therefore, predictive biomarkers are needed to guide the selection of patients who are likely to respond favorably to immunotherapeutic agents, thereby alleviating the financial burden on patients and improving their survival. Programmed death-ligand 1 (PD-L1) expression, deficiency in mismatch repair proteins, or microsatellite instability are commonly used biomarkers for selecting patients suitable for immunotherapy. The efficacy and duration of response to ICIs are influenced by multiple biological factors in addition to PD-L1 expression. Previous studies have shown that an increase in peripheral blood eosinophils indicates a better prognosis in patients with non-small cell lung cancer (NSCLC) receiving immunotherapy.²¹ Furthermore, a low peripheral blood neutrophil-to-lymphocyte ratio (NLR) and low lactate dehydrogenase (LDH) levels were associated with longer PFS (P=0.049 and 0.046, respectively) and OS (P=0.007 and 0.031, respectively) in patients with advanced NSCLC receiving programmed cell death protein-1 (PD-1) inhibitor therapy.²² High levels of C-reactive protein (CRP) at baseline were associated with worse PFS (P=0.005) and OS (P=0.00004) in cancer patients receiving immunotherapy, suggesting a potential predictive value for anti-PD-1/PD-L1 agents.²³ In a phase II trial of pembrolizumab and vorinostat in patients with recurrent/metastatic head and neck squamous cell carcinoma and salivary gland cancer, high baseline levels of NLR and neutrophils, as well as low baseline levels of lymphocytes and T helper cells, were strongly associated with worse OS and PFS.²⁴ Currently, no reliable biomarkers have been identified to predict the response to ICIs therapy in patients with HCC. Evidence supporting the relationship between PD-L1 expression and the efficacy of ICIs mostly comes from clinical trials, which vary considerably in their designs and treatment regimens. Most patients with cancer do not respond to or develop resistance to PD-1/PD-L1 blockade immunotherapy. Tumor cell PD-L1 expression correlates with objective responses to immunotherapy in tumors such as melanoma and NSCLC, but PD-L1 positivity alone is insufficient for patient stratification, because some patients with negative PD-L1 expression also respond favorably to immunotherapy.^25,26 For example, in the CheckMate-040 clinical study, regardless of PD-L1 expression levels, HCC patients treated with nivolumab showed improved 9-month OS and PFS.²⁷ Moreover, due to the use of different PD-L1 clones, staining platforms, and PD-L1 expression scoring methods, the reliability of detecting tumor PD-L1 expression by immunohistochemistry remains controversial.^28,29 Therefore, the role of PD-L1 expression in predicting the efficacy of immunotherapy for HCC is debatable The low tumor mutational burden and microsatellite instability levels in HCC limit their utility as predictive factors for response to immunotherapy.^30,31 Therefore, further exploration of reliable biomarkers is needed to select patients with HCC who may respond to immunotherapy. Studies on peripheral hematological parameters associated with the efficacy of PD‑1/PD‑L1 inhibitors in patients with HCC are limited. This study aimed to evaluate the association of clinical characteristics and peripheral hematological parameters with the efficacy and prognosis of patients with HCC receiving PD‑1/PD‑L1 inhibitor therapy, and to explore factors associated with poor prognosis in patients with HCC undergoing immunotherapy.

Methods

Research Subjects

This study retrospectively collected the demographic and clinical baseline data of patients with HCC who received PD-1/PD-L1 inhibitor therapy between January 2019 and October 2022. The inclusion criteria were as follows: (1) age ≥18 years, (2) diagnosis of HCC, (3) receipt of at least one cycle of PD-1/PD-L1 inhibitor therapy, and (4) presence of at least one measurable lesion on imaging. The exclusion criteria were as follows: (1) presence of another primary malignancy elsewhere; (2) presence of an infectious disease within one week prior to the first dose of immunotherapy; (3) history of any of the following: solid organ transplantation, bone marrow transplantation, primary hematological disease, autoimmune deficiency, or autoimmune disease; (4) use of corticosteroids within one month prior to the first immunotherapy; and (5) loss to follow-up or incomplete medical records before HCC progression.

This study collected relevant clinical data and factors potentially affecting prognosis from electronic medical records, which were categorized into the following five groups: (1) patient baseline characteristics: age, sex, body mass index (BMI), Eastern Cooperative Oncology Group (ECOG) performance status, and Child-Pugh class; (2) tumor-related characteristics: HCC clinical stage, vascular invasion, distant metastasis, number and sites of metastatic organs, follow-up outcomes, and time from initiation of PD-1/PD-L1 inhibitor therapy to HCC progression or death; (3) laboratory findings: selected laboratory parameters measured in venous blood drawn within one week prior to the first immunotherapy, such as white blood cell count (WBC), red blood cell count (RBC), platelet count (PLT), hemoglobin (Hb), absolute monocyte count (AMC), absolute neutrophil count (ANC), absolute lymphocyte count (ALC), alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), C-reactive protein (CRP), gamma-glutamyl transferase (GGT), albumin (ALB), and total bilirubin (TBIL); (4) imaging features: examinations such as computed tomography (CT) or magnetic resonance imaging (MRI) performed at 4–8 weeks after immunotherapy or at any time based on clinical evaluation by physicians; (5) treatment details: line of therapy, presence and regimen of concomitant targeted therapy, presence of concomitant local therapy, presence and treatment status of viral hepatitis, and best treatment response.

Treatment Regimens

The patient treatment regimens included: camrelizumab, 200 mg intravenous infusion every 3 weeks; nivolumab, 480 mg intravenous infusion every 4 weeks; sintilimab, 200 mg intravenous infusion every 3 weeks; toripalimab, 240 mg intravenous infusion every 3 weeks; envafolimab, 300 mg subcutaneous injection every 3 weeks; atezolizumab, 1200 mg intravenous infusion every 3 weeks; tislelizumab, 200 mg intravenous infusion every 3 weeks; and pembrolizumab, 200 mg intravenous infusion every 3 weeks. Specific treatment regimens could be combined with chemotherapy, anti-angiogenic agents, or other targeted drugs based on the physician’s assessment, or PD-1/PD-L1 inhibitors could be used alone. The actual drug dosage and administration intervals were adjusted according to patient tolerance and physician evaluation.

Prognostic Assessment

Efficacy was evaluated according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Efficacy outcomes were classified into four categories: complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The objective response rate (ORR) and disease control rate (DCR) were used as efficacy endpoints, calculated using the following formula: ORR = (CR + PR)/(CR + PR + SD + PD) × 100%, and DCR = (CR + PR + SD)/(CR + PR + SD + PD) × 100%. PFS was defined as the time from the first use of PD-1/PD-L1 inhibitors to tumor progression or patient death. We followed up patients through electronic medical records until October 31, 2022, and 101 patients were followed up and their disease status was confirmed.

Statistical Analysis

Statistical analyses were performed using SPSS Version 26.0 and X-tile software. The cutoff values of continuous variables such as collected peripheral blood test parameters were determined using X-tile software, and the continuous variables were converted into binary variables based on these cutoff values.³² The Pearson χ2 test or Fisher’s exact test was used for intergroup comparisons of categorical variables. Univariate analysis was performed using the Cox regression model. Clinical characteristics and laboratory parameters (P < 0.1 in the univariate analysis were subjected to the Log rank test, and Kaplan-Meier (K-M) curves were plotted. Subsequently, multivariate analysis was performed using the Cox regression model, and statistical significance was set at P < 0.05. Independent risk factors were also identified.

Results

Clinical Characteristics of the Patients

A total of 101 patients with hepatocellular carcinoma (HCC) were enrolled in this study, including 83 males (82.18%) and 18 females (17.82%), with a median age of 58 years (range, 31–85 years). A total of 29 patients had a BMI >24, 63 patients had a BMI between 18.5 and 24, and 9 patients had a BMI <18.5. Regarding ECOG performance status, 35 patients scored 0, 55 patients scored 1, and 11 patients scored 2. For Child-Pugh class, 90 patients (89.1%) were class A and 11 patients (10.9%) were class B. Most patients were positive for hepatitis B surface antigen (HBsAg) (81.2%), and among these HBsAg-positive patients, 96.3% received antiviral therapy. This study adopted the China Liver Cancer Staging (CNLC) system, among which 66 patients (65.3%) had stage III/IV disease and 35 patients (34.7%) had other stages. The detailed baseline characteristics of the patients are presented in Table 1. All patients received PD-1/PD-L1 inhibitor immunotherapy, with 74 patients (73.3%) receiving immunotherapy-based first-line therapy and 27 patients (26.7%) receiving immunotherapy-based second-line or later therapies. A total of 83 patients (82.2%) received concomitant antitumor agents, all of which were targeted agents, while 18 patients (17.8%) received immunotherapy as monotherapy. By the end of the follow-up, disease progression or death was observed in 69 patients, with an ORR of 8.91% and a DCR of 80.20%.

Table 1.

Baseline Characteristics of the Patients

Characteristics	N, Percentage
Age
< 60 years old	55 (54.5%)
≥ 60 years old	46 (45.5%)
Gender
Male	83 (82.2%)
Female	18 (17.8%)
BMI
<18.5	9 (8.9%)
18.5~24	63 (62.4%)
> 24	29 (28.7%)
CNLC
III/IV	66 (65.3%)
Other	35 (34.7%)
ECOG-PS
0	35 (34.7%)
1	55 (54.4%)
2	11 (10.9%)
Child-Pugh
A	90 (89.1%)
B	11 (10.9%)
Hepatitis B
Yes	82 (81.2%)
No	19 (18.8%)
Antiviral therapy
Yes	79 (78.2%)
No	22 (21.8%)
Line of treatment
1	74 (73.3%)
≥ 2	27 (26.7%)
Immunotherapy agents
Sintilimab	20 (19.8%)
Tislelizumab	16 (15.8%)
Toripalimab	13 (12.9%)
Camrelizumab	38 (37.6%)
Atezolizumab	5 (5%)
Envafolimab	5 (5%)
Other	4 (4%)
Concomitant antitumor agents
Yes	83 (82.2%)
No	18 (17.8%)
Concomitant targeted agents
No	18 (17.8%)
Apatinib	21 (20.8%)
Lenvatinib	33 (32.7%)
Regorafenib	9 (8.9%)
Other	20 (19.8%)
Best response
PR/CR	9 (8.9%)
SD	72 (71.3%)
PD	20 (19.8%)
Concomitant local therapy
Yes	20 (19.8%)
No	81 (80.2%)
Distant metastasis
Yes	43 (42.6%)
No	58 (57.4%)
Lung metastasis	17 (16.8%)
Bone metastasis	6 (5.9%)
Omental metastasis	8 (7.9%)
Adrenal metastasis	5 (5.0%)
Other distant sites	27 (26.7%)
Number of metastatic organs
None	58 (57.4%)
1	13 (12.9%)
≥ 2	30 (29.7%)
WBC
≥ 5	38 (37.62%)
< 5	63 (62.38%)
RBC
≥ 4.19	53 (52.5%)
< 4.19	48 (47.5%)
Hb
≥ 140	37 (36.6%)
< 140	64 (63.4%)
PLT
≥ 152	36 (35.6%)
< 152	65 (64.4%)
ANC
≥ 2.5	68 (67.3%)
< 2.5	33 (32.7%)
ALC
≥ 1.39	31 (30.7%)
< 1.39	70 (69.3%)
AMC
≥ 0.49	22 (21.8%)
< 0.49	79 (78.2%)
ALT
≥ 69	12 (11.9%)
< 69	89 (88.1%)
AST
≥ 26	83 (82.2%)
< 26	18 (17.8%)
ALP
≥ 215	17 (16.8%)
< 215	84 (83.2%)
GGT
≥ 246	11 (10.9%)
< 246	90 (89.1%)
CRP
≥ 14.6	28 (27.7%)
< 14.6	73 (72.3%)
ALB
≥ 37.7	49 (48.5%)
< 37.7	52 (51.5%)
TBIL
≥ 24.5	20 (19.8%)
< 24.5	81 (80.2%)

Abbreviations: BMI, Body Mass Index; CNLC, China Liver Cancer Staging; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; CR, Complete Response; PR, Partial Response; SD, Stable Disease; PD, Progressive Disease; WBC, White Blood Cell Count; RBC, Red Blood Cell Count; Hb, Hemoglobin; PLT, Platelet Count; ANC, Absolute Neutrophil Count; ALC, Absolute Lymphocyte Count; AMC, Absolute Monocyte Count; ALT, Alanine Aminotransferase; AST, Aspartate Aminotransferase; ALP, Alkaline Phosphatase; GGT, Gamma-glutamyl Transferase; CRP, C-reactive protein; ALB, Albumin; TBIL, Total Bilirubin.

Optimal Cutoff

Based on patient progression-free survival time and survival status, the optimal cutoff values for peripheral blood laboratory parameters were determined using the X-tile software. The cutoff values were as follows: WBC 5 × 10⁹/L, RBC 4.19 × 10¹²/L, Hb 140 g/L, PLT 152 × 10⁹/L, ANC 2.5 × 10⁹/L, ALC 1.39 × 10⁹/L, AMC 0.49 × 10⁹/L, ALT 69 U/L, AST 26 U/L, ALP 215 U/L, GGT 246 U/L, CRP 14.6 mg/L, ALB 37.7 g/L, and TBIL 24.5 μmol/L. These cutoff values were used to convert continuous variables into binary variables.

Assessment

Based on patient responses to immunotherapy, by the end of follow-up, 2 patients achieved CR, 7 patients achieved PR, 20 patients were assessed as having PD, and 72 patients were assessed as having SD. According to the above formulas for ORR and DCR, ORR and DCR were calculated for each clinical characteristic and peripheral blood parameter subgroup. Significant differences in ORR were observed among the patients stratified by BMI (P=0.008). Significant differences in DCR were observed among patients stratified by Child-Pugh class (P=0.039), line of therapy (P=0.009), distant metastasis (P=0.024), bone metastasis (P=0.013), ALP level (P=0.039), and TBIL level (P=0.024) (Table 2).

Table 2.

Response to Treatment

Clinical Indices	ORR	P	DCR	P
Age		0.504		0.119
< 60 years old	10.91%		74.55%
≥ 60 years old	6.52%		86.96%
Gender		0.198		0.751
Male	7.23%		80.72%
Female	16.67%		77.78%
BMI		0.008		0.611
<18.5	33.33%		66.67%
18.5~24	6.35%		79.37%
> 24	6.90%		86.21%
CNLC		1.000		0.311
III/IV	9.09%		77.27%
Other	8.57%		85.71%
ECOG-PS		0.248		0.166
0	2.86%		88.57%
1	12.73%		78.18%
2	9.09%		63.64%
Child-Pugh		1.000		0.039
A	8.89%		83.33%
B	9.09%		54.55%
Hepatitis B		0.364		1.000
Yes	7.32%		80.49%
No	15.79%		78.95%
Antiviral therapy		0.405		0.367
Yes	7.59%		82.28%
No	13.64%		72.73%
Line of treatment		1.000		0.009
1	9.46%		86.49%
≥ 2	7.41%		62.96%
Immunotherapy agents		0.897		0.925
Sintilimab	10.00%		80.00%
Tislelizumab	6.25%		87.50%
Toripalimab	7.69%		69.23%
Camrelizumab	10.53%		81.58%
Atezolizumab	0%		80%
Envafolimab	0%		80%
Other	25.00%		75.00%
Concomitant antitumor agents		0.198		1.000
Yes	7.23%		79.52%
No	16.67%		83.33%
Concomitant targeted agents		0.656		0.798
No	16.67%		83.33%
Apatinib	9.52%		76.19%
Lenvatinib	12.12%		81.82%
Regorafenib	0%		55.55%
Other	0%		90.00%
Concomitant local therapy		0.684		0.348
Yes	5.00%		90.00%
No	9.88%		77.78%
Distant metastasis		0.729		0.024
Yes	6.98%		69.77%
No	10.34%		87.93%
Lung metastasis		0.350		0.319
Yes	0%		70.59%
No	10.71%		82.14%
Bone metastasis		0.437		0.013
Yes	16.67%		33.33%
No	8.42%		83.16%
Omental metastasis		0.539		0.192
Yes	12.50%		62.50%
No	8.60%		81.72%
Adrenal metastasis		0.062		0.256
Yes	40.00%		60.00%
No	7.29%		81.25%
Other distant sites		0.725		0.134
Yes	11.11%		70.37%
No	9.46%		83.78%
Number of metastatic organs		0.775		0.057
None	10.34%		87.93%
1	0%		76.92%
≥ 2	10.00%		66.67%
WBC		0.291		0.193
≥ 5	13.16%		86.84%
< 5	6.35%		76.19%
RBC		0.302		0.805
≥ 4.19	5.66%		81.13%
< 4.19	12.50%		79.17%
Hb		0.480		0.727
≥ 140	5.41%		78.38%
< 140	10.94%		81.25%
PLT		0.879		0.103
≥ 152	8.33%		88.89%
< 152	9.23%		75.38%
ANC		0.265		0.065
≥ 2.5	11.76%		85.29%
< 2.5	3.03%		69.70%
ALC		0.268		0.538
≥ 1.39	3.23%		83.87%
< 1.39	11.43%		78.57%
AMC		1.000		0.367
≥ 0.49	9.09%		72.73%
< 0.49	8.86%		82.28%
ALT		1.000		0.451
≥ 69	8.33%		91.67%
< 69	8.99%		78.65%
AST		0.051		0.114
≥ 26	6.02%		77.11%
< 26	22.22%		94.44%
ALP		0.645		0.039
≥ 215	11.76%		58.82%
< 215	8.33%		84.52%
GGT		1.000		1.000
≥ 246	9.09%		81.82%
< 246	8.89%		80.00%
CRP		0.689		0.417
≥ 14.6	10.71%		75.00%
< 14.6	8.22%		82.19%
ALB		1.000		0.064
≥ 37.7	8.16%		87.76%
< 37.7	9.62%		73.08%
TBIL		0.683		0.024
≥ 24.5	5.00%		60.00%
< 24.5	9.88%		85.19%

Abbreviations: ORR, objective response rate; DCR, disease control rate; BMI, body mass index; CNLC, China Liver Cancer Staging; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; PLT, platelet count; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; AMC, absolute monocyte count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; CRP, C-reactive protein; ALB, albumin; TBIL, total bilirubin.

Survival Analysis

Univariate Cox analysis identified lung metastasis (P=0.009), number of metastatic organs (P=0.045), best response (P<0.001), Hb (P=0.042), ANC (P=0.044), ALT (P=0.025), AST (P=0.014), ALP (P=0.001), GGT (P=0.016), CRP (P=0.017), and TBIL (P=0.001) as potential prognostic factors for PFS in HCC patients receiving immunotherapy with PD-1/PD-L1 inhibitors. In contrast, age, sex, BMI, clinical stage, ECOG PS, Child-Pugh class, hepatitis B, antiviral therapy, line of therapy, type of immune checkpoint inhibitor, concomitant antitumor therapy, concomitant targeted therapy, local therapy, distant metastasis, bone metastasis, omental metastasis, adrenal metastasis, other distant sites, WBC, RBC, PLT, ALC, AMC, and ALB were not significantly different (Table 3).

Table 3.

Univariate Analysis of PFS

Clinical Indices	HR (95% CI)	p
Age (≥60 vs<60)	0.770 (0.472–1.225)	0.294
Gender (Female vs Male)	0.840 (0.649–1.876)	0.615
BMI		0.596
<18.5	Ref.
18.5~24	0.727 (0.305–1.732)	0.472
> 24	0.919 (0.368–2.293)	0.856
CNLC (III/IV vs Other)	1.286 (0.768–2.154)	0.339
ECOG-PS		0.767
0	Ref.
1	0.880 (0.518–1.496)	0.637
2	0.755 (0.345–1.654)	0.483
Child-Pugh (B vs A)	1.254 (0.619–2.539)	0.529
Hepatitis B (Yes vs No)	0.867 (0.478–1.571)	0.638
Antiviral therapy (Yes vs No)	1.384 (0.793–2.416)	0.252
Line of treatment (≥2vs1)	1.478 (0.882–2.476)	0.138
Immunotherapy agents		0.830
Other	Ref.
Sintilimab	1.504 (0.318–7.120)	0.607
Tislelizumab	2.070 (0.414–10.346)	0.375
Toripalimab	2.416 (0.490–11.902)	0.278
Camrelizumab	2.466 (0.532–11.436)	0.249
Atezolizumab	2.110 (0.271–16.413)	0.476
Envafolimab	2.128 (0.178–25.407)	0.550
Concomitant antitumor agents (Yes vs No)	1.628 (0.888–2.985)	0.115
Concomitant targeted agents		0.500
No	Ref.
Apatinib	1.678 (0.825–3.412)	0.153
Lenvatinib	1.621 (0.798–3.293)	0.182
Regorafenib	2.240 (0.837–5.995)	0.108
Other	1.374 (0.587–3.214)	0.464
Concomitant local therapy (Yes vs No)	0.669 (0.350–1.282)	0.226
Distant metastasis (Yes vs No)	1.461 (0.909–2.349)	0.117
Lung metastasis (Yes vs No)	2.191 (1.216–3.949)	0.009
Bone metastasis (Yes vs No)	2.423 (0.962–6.100)	0.060
Omental metastasis (Yes vs No)	1.531 (0.694–3.377)	0.292
Adrenal metastasis (Yes vs No)	0.727 (0.263–2.015)	0.540
Other distant sites (Yes vs No)	0.998 (0.602–1.656)	0.995
Number of metastatic organs		0.125
None	Ref.
1	2.024 (1.017–4.030)	0.045
≥2	1.301 (0.772–2.195)	0.323
Best response		<0.001
PR/CR	Ref.
SD	2.192 (0.859–5.592)	0.100
PD	35.439 (11.179–112.345)	<0.001
WBC (≥5 vs <5)	0.773 (0.472–1.268)	0.308
RBC (≥4.19 vs <4.19)	1.461 (0.897–2.379)	0.128
Hb (≥140 vs <140)	1.655 (1.018–2.690)	0.042
PLT (≥152 vs <152)	0.714 (0.426–1.195)	0.199
ANC (≥2.5 vs <2.5)	0.597 (0.361–0.986)	0.044
ALC (≥1.39 vs <1.39)	1.418 (0.866–2.324)	0.165
AMC (≥0.49 vs <0.49)	1.554 (0.903–2.676)	0.112
ALT (≥69 vs <69)	2.242 (1.109–4.532)	0.025
AST (≥26 vs <26)	2.202 (1.170–4.147)	0.014
ALP (≥215 vs <215)	3.394 (1.704–6.758)	0.001
GGT (≥246 vs <246)	2.697 (1.202–6.052)	0.016
CRP (≥14.6 vs <14.6)	1.977 (1.130–3.459)	0.017
ALB (≥37.7 vs <37.7)	0.680 (0.420–1.101)	0.117
TBIL (≥24.5 vs <24.5)	2.580 (1.443–4.610)	0.001

Abbreviations: Ref., reference; BMI, body mass index; CNLC, China Liver Cancer Staging; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; PLT, platelet count; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; AMC, absolute monocyte count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; CRP, C-reactive protein; ALB, albumin; TBIL, total bilirubin.

Univariate Cox analysis identified 12 variables with P < 0.1, including lung metastasis, bone metastasis, number of metastatic organs, best response, Hb, ANC, ALT, AST, ALP, GGT, CRP, and TBIL. Kaplan-Meier (K‑M) curves were plotted for all study factors (P < 0.1 to perform survival analysis. Survival curves indicated that patients without metastatic organs had longer PFS than those with one metastatic organ (median PFS [mPFS] 209 days vs 114 days, χ²=4.453, P=0.035) (Figure 1A). This study further showed that patients with lung metastasis had shorter PFS than those without lung metastasis (mPFS 114 days vs 206 days, χ²=7.177, P=0.007) (Figure 1B). The mPFS was 62 days in patients with bone metastasis and 188 days in those without bone metastasis, with no statistically significant difference between the two groups (P = 0.052) (Figure 1C). There was a significant difference in the mPFS between patients with the best response of PR/CR and those with the best response of PD (mPFS 409 days vs 63 days, χ²=20.504, P<0.001). Patients with the best response of SD had longer mPFS than those with the best response of PD (mPFS 223 days vs 63 days, χ²=78.991, P<0.001) (Figure 1D). Furthermore, the mPFS was 215 days in patients with Hb <140 and 145 days in those with Hb ≥140, with a statistically significant difference between the two groups (P=0.040) (Figure 1E). In contrast, a study in patients with non-small cell lung cancer (NSCLC) receiving immunotherapy found that patients with normal baseline Hb levels had significantly longer PFS than those with decreased Hb counts (<110 g/L) (mPFS 10.0 vs 4.0 months, P=0.001).³³ This is in contrast to our results. On one hand, the patients with Hb <140 g/L in our study included some individuals with normal baseline Hb levels (male Hb≥120 g/L, female Hb ≥110 g/L) as well as those with decreased baseline Hb levels, which differs from the design of other studies, and further investigation is needed regarding the grouping of these patients. However, the limited number of cases included in our study may account for this discrepancy in the findings. Patients in the ANC≥2.5×10⁹/L group had significantly longer PFS than those in the ANC <2.5×10⁹/L group (mPFS 215 days vs 132 days, χ²=4.148, P=0.042) (Figure 1F). Statistically significant differences in PFS were also observed between the baseline ALT ≥69 U/L and ALT <69 U/L groups (mPFS 138 days vs 209 days, χ²=5.348, P=0.021), and between the AST ≥26 U/L and AST <26 U/L groups (mPFS 145 days vs 352 days, χ²=6.275, P=0.012) (Figure 1G and H). Patients with ALP < 215 U/L had significantly longer PFS than those with ALP ≥ 215 U/L (mPFS 206 days vs 83 days, χ²=13.646, P < 0.001; multivariate Cox regression: HR = 0.246, P = 0.014 for ALP < 215 U/L vs ALP ≥ 215 U/L) (Figure 1I). Elevated GGT, a nonspecific indicator that can be seen in biliary tract diseases, hepatic malignancies, drug‑induced liver injury, fatty liver disease, autoimmune liver disease, etc, showed a significant difference in PFS between patients with GGT ≥246 U/L and those with GGT <246 U/L (mPFS 101 days vs 204 days, χ²=6.286, P=0.012) (Figure 1J). Patients with serum TBIL elevation of grade 2 or higher should discontinue immunotherapy, and their serum transaminase and TBIL levels should be monitored to mitigate immune-related hepatotoxicity. This study found that patients in the TBIL <24.5 μmol/L group had significantly better PFS than those in the TBIL ≥24.5 μmol/L group (mPFS 223 days vs 101 days, χ²=11.003, P=0.001) (Figure 1K). The reference range for CRP level at our center was <6.0 mg/L. CRP levels may be elevated to varying degrees in patients with malignancies, tissue injuries, infections, or autoimmune diseases. Our results showed that the mPFS was 92 days in the CRP ≥14.6 mg/L group and 206 days in the CRP <14.6 mg/L group, with a statistically significant difference (χ²=5.926, P=0.015) (Figure 1L).

Figure 1.

Kaplan-Meier curves for progression-free survival (PFS) of 12 variables identified by univariate Cox regression analysis. Each subpanel represents the PFS comparison between groups stratified as follows: (A) Number of metastatic organs (1 vs None; 1 vs ≥2; ≥2 vs None); (B) Lung metastasis (presence vs absence); (C) Bone metastasis (presence vs absence); (D) Best response (PR/CR vs SD; PR/CR vs PD; SD vs PD); (E) Hemoglobin (Hb); (F) Absolute neutrophil count (ANC); (G) Alanine aminotransferase (ALT); (H) Aspartate aminotransferase (AST); (I) Alkaline phosphatase (ALP); (J) Gamma-glutamyl transferase (GGT); (K) Total bilirubin (TBIL); (L) C-reactive protein (CRP). For continuous variables (E–L), patients were stratified by the cutoff value. Differences in PFS between groups were compared using the Log rank test, and P < 0.05 was considered statistically significant.

Abbreviations: PFS, progression-free survival; Hb, hemoglobin; ANC, absolute neutrophil count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; TBIL, total bilirubin; CRP, C-reactive protein.

Multivariate Cox regression analysis was performed on clinical indices (P < 0.1 in the univariate analysis (lung metastasis, bone metastasis, number of metastatic organs, best response, Hb, ANC, ALT, AST, ALP, GGT, CRP, and TBIL). Best response was included in the multivariate Cox regression model as a categorical variable with three levels (PR/CR, SD, and PD), with PR/CR as the reference group. The results showed that lung metastasis (HR = 2.823, P = 0.011), best response (HR = 51.634, P < 0.001), Hb (HR = 2.204, P = 0.009), ANC (HR = 0.528, P = 0.039), ALP (HR = 0.246, P = 0.014), GGT (HR = 8.100, P = 0.002), and CRP (HR = 2.316, P = 0.017) were independent prognostic factors for PFS in patients with HCC receiving immunotherapy (Table 4).

Table 4.

Multivariate Analysis of PFS

Clinical Indices	HR (95% CI)	p
Lung metastasis (Yes vs No)	2.823 (1.267–6.291)	0.011
Bone metastasis (Yes vs No)	1.881 (0.570–6.204)	0.300
Number of metastatic organs (1 vs None)	0.968 (0.383–2.448)	0.946
Best response (PD vs PR/CR)	51.634 (12.228–218.035)	<0.001
Hb (≥140 vs <140)	2.204 (1.215–4.001)	0.009
ANC (≥2.5 vs <2.5)	0.528 (0.287–0.969)	0.039
ALT (≥69 vs <69)	1.573 (0.551–4.493)	0.398
AST (≥26 vs <26)	1.364 (0.608–3.063)	0.451
ALP (≥215 vs <215)	0.246 (0.080–0.752)	0.014
GGT (≥246 vs <246)	8.100 (2.196–29.877)	0.002
CRP (≥14.6 vs <14.6)	2.316 (1.165–4.606)	0.017
TBIL (≥24.5 vs <24.5)	1.609 (0.695–3.727)	0.267

Abbreviations: PD, progressive disease; PR, partial response; CR, complete response; SD, Stable Disease; Hb, hemoglobin; ANC, absolute neutrophil count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; CRP, C-reactive protein; TBIL, total bilirubin. Best response coded as PR/CR (reference), SD, PD.

Discussion

HCC is one of the most important global health issues, and most patients eventually require systemic therapy. The TME is a dynamic milieu composed of tumor cells, stromal cells, immune cells, cytokines, the extracellular matrix, and other components. As a major modality of systemic therapy, immunotherapy exerts its antitumor effects by modifying the TME. In recent years, PD‑1/PD‑L1 inhibitors have emerged as promising agents in the immunotherapy landscape for HCC, offering patients with additional therapeutic options. However, atypical responses following ICIs therapy, including pseudoprogression, hyperprogression, delayed response, and dissociated response, may hinder the selection of patients most likely to benefit from immunotherapy. Furthermore, resistance and toxicity arising during immunotherapy can affect long-term survival. PD‑1/PD‑L1 inhibitors are not effective for all patients, and effective prognostic biomarkers for immunotherapy are still lacking.³⁴ In the CheckMate‑040 trial, patients with HCC who received immunotherapy achieved CR or PR regardless of PD‑L1 expression levels.³⁵ Therefore, while clinical studies involving various immunotherapeutic agents are increasing rapidly, the exploration of accurate and effective prognostic biomarkers for immunotherapy may facilitate personalized treatment. However, correctly interpreting these findings requires careful attention to the distinction between prognostic and predictive biomarkers. Prognostic markers are associated with clinical outcomes independently of treatment, whereas predictive markers identify patients likely to derive differential benefit from a specific therapy. Given that this was a single‑arm retrospective study in which all patients received PD‑1/PD‑L1 inhibitor‑based regimens, the factors identified in our multivariate analysis should be interpreted as prognostic factors for progression‑free survival (PFS) in this treatment context, rather than as predictive biomarkers of immunotherapy response. Definitive confirmation of predictive value would require a comparative design with a non‑immunotherapy control arm, which was beyond the scope of the present study.

Real‑world evidence has increasingly emphasized the importance of host factors in shaping outcomes during ICI therapy. Wang et al reported that patients with liver cancer had the second‑highest incidence of cardiovascular adverse events (33.1%) across all cancer types in a cohort of 9,541 ICI‑treated patients.³⁶ This study suggests that in the investigation of HCC patients receiving immunotherapy, attention should also be paid to their immune‑related adverse events. Furthermore, Wang et al identified the rs16957301 variant in the PCCA gene as a risk genotype specifically associated with ICI‑related acute kidney injury (OR = 2.53, P = 0.047).³⁷ This finding suggests that incorporating genetic markers alongside clinical and hematological parameters may enhance the prognostic framework for HCC patients receiving immunotherapy.

Previous studies have demonstrated a relationship between baseline clinical indices and prognosis in patients receiving immune checkpoint inhibitor therapy. Compared with patients with gestational trophoblastic neoplasia without lung metastases, those with lung metastases had a significantly higher rate of drug resistance (P = 0.019) and a relatively higher mortality rate (P = 0.031).³⁸ Therefore, the presence of lung metastasis before treatment may be associated with an increased risk of tumor progression and recurrence. Our study also indicated that patients with HCC and lung metastasis who received immunotherapy had significantly shorter PFS. High relative eosinophil count (≥1.5%) and relative lymphocyte count (≥17.5%), as well as LDH elevation ≤2.5 times the upper limit of normal, were independent baseline characteristics associated with favorable OS in patients with melanoma treated with pembrolizumab.³⁹ Serum levels of CRP and AFP have also been demonstrated to be independent risk factors for PFS in patients with hepatocellular carcinoma treated with PD‑1 inhibitors (P < 0.05).⁴⁰ Furthermore, a study reported that in patients with stage IV NSCLC receiving PD‑1 inhibitor therapy, liver metastasis (P = 0.01), neutrophil‑to‑lymphocyte ratio (P = 0.001), and lymphocyte‑to‑monocyte ratio (P = 0.006) were independent prognostic indicators for OS.⁴¹ A large‑scale pan‑cancer cohort study of 1,479 patients treated with immune checkpoint blockade therapy found that higher baseline albumin levels were significantly associated with improved OS, PFS, and ORR (P < 0.001).⁴² Based on these findings, through analysis of baseline clinical indices from 101 patients with HCC treated with PD‑1/PD‑L1 inhibitors, the present study identified lung metastasis (P = 0.011), best response (P < 0.001), Hb (P = 0.009), ANC (P = 0.039), ALP (P = 0.014), GGT (P = 0.002), and CRP (P = 0.017) as independent prognostic factors for PFS in patients with HCC.

Neutrophils and CRP play important roles in the immune microenvironment. Immune activation is a dynamic process, and inflammatory biomarkers, which can accurately reflect tumor–immune interactions, represent promising prognostic tools. A Among patients with non‑small cell lung cancer treated with PD‑1/PD‑L1 inhibitors, baseline levels of procalcitonin, a reliable marker of bacterial infection, below a predetermined cutoff were associated with better prognosis (P < 0.05).⁴³ Higher CRP is associated with a tumor immunosuppressive microenvironment in renal cell carcinoma, leading to worse prognosis.⁴⁴ CRP can inhibit the proliferation and activation of CD4+ and CD8+ T cells in patients with melanoma, downregulate the expression of costimulatory molecules on mature dendritic cells in a dose‑dependent manner, and suppress the expansion of MART‑1‑specific CD8+ T cells.⁴⁵ Yoshida et al showed that elevated CRP were significantly associated with shorter survival in patients with melanoma treated with nivolumab and ipilimumab.⁴⁵ The present study analyzed the relationship between CRP and prognosis in patients with HCC receiving PD‑1/PD‑L1 inhibitors and found that patients with higher baseline CRP had significantly shorter PFS compared with those with lower CRP (P = 0.017). Neutrophils consist of distinct subpopulations with either pro-tumor or anti-tumor roles. Tumor‑associated neutrophils can exert marked antitumor effects by strongly suppressing a subset of T cells that produce interleukin‑17, which promotes tumor growth and metastasis.⁴⁶ Studies have shown that neutrophils with antitumor effects exhibit pronounced PD‑L1 expression.⁴⁷ Neutrophil extracellular traps promote the activation of naïve CD4+ T cells through metabolic reprogramming, and activated regulatory T cells can suppress immune surveillance, thereby promoting tumor progression.⁴⁸ Due to the plasticity, heterogeneity, and diversity of neutrophils, the relationship between neutrophils and cancer prognosis remains unclear. In the present study, patients with a higher ANC had longer PFS than those with a lower ANC (P = 0.042).

LDH and PLT may influence patient prognosis by regulating tumor angiogenesis. LDH is a glycolytic enzyme that converts pyruvate to lactate under aerobic conditions.⁴⁹ Most tumor cells rely on aerobic glycolysis to generate the energy they require, and LDH expression is typically increased in tumor cells to facilitate the metabolic switch to aerobic glycolysis.⁵⁰ LDH promotes tumor angiogenesis and enhances invasiveness through aberrant activation of hypoxia‑inducible factor 1, and also contributes to resistance to therapy.^51,52 Studies have shown that HCC patients with high plasma LDH had a lower mOS than those with low LDH (P < 0.001).⁵³ However, findings regarding the relationship between LDH and the degree of benefit in HCC patients treated with sorafenib have been inconsistent.^54,55 PLT promotes wound healing and maintains vascular integrity by adhering, activating, and aggregating at sites of vascular injury. Accumulating evidence indicates that PLT plays a critical role in promoting inflammation, as well as tumor initiation and progression.⁵⁶ PLT can promote tumor progression by modulating angiogenesis, and P-selectin expressed on platelets interacts with tumor cells to stimulate immune evasion and angiogenesis.^57,58 Compared with patients with normal PLT, those with elevated baseline PLT in NSCLC and locally advanced pancreatic cancer had significantly worse prognosis.^59,60 Elevated baseline PLT was an independent predictor of poor response to TACE (P = 0.001) and shorter time to disease progression (HR = 3.598; 95% CI 2.570‑5.036, P < 0.001) in HCC.⁶¹ In contrast, our study found no statistically significant difference in PFS among HCC patients receiving immunotherapy with different baseline PLT.

Elevated baseline AFP and GGT levels are associated with tumor vascular invasion, thereby enhancing tumor aggressiveness. AFP is a glycoprotein that remains at very low levels under normal physiological conditions. Elevated serum AFP levels are observed in 60%–70% of patients with HCC, and AFP is one of the main biomarkers for early screening of HCC.⁶² Persistently elevated AFP levels are associated with tumor vascular invasion and poor differentiation, leading to increased aggressiveness of HCC.^62,63 Elevated baseline AFP has been shown to be a poor prognostic factor for OS in patients with HCC.⁶⁴ HCC patients with high AFP levels are more likely to have bilobar liver involvement and portal vein thrombosis.⁶⁵ Because baseline LDH and AFP values were missing for a large proportion of patients in the present study, these two factors were not analyzed. Future studies should explore the relationship between baseline levels of LDH and AFP and prognosis in HCC patients receiving immunotherapy. GGT is a membrane‑bound enzyme involved in the synthesis and degradation of glutathione.⁶⁶ GGT can induce DNA damage and promote tumor microvascular invasion and epithelial‑mesenchymal transition through various signaling pathways, and can also block the entry of chemotherapeutic drugs into tumor cells, thereby leading to drug resistance.⁶⁷ Furthermore, GGT can induce CpG island methylation in certain regions of the genome, contributing to drug resistance and recurrence in HCC.⁶⁸ Lee et al demonstrated that repeatedly elevated serum GGT are associated with the risk of developing respiratory system tumors.⁶⁹ Elevated baseline GGT is associated with reduced OS in patients with metastatic melanoma receiving ICI therapy.⁷⁰ Ishiyama et al found in multivariate analysis that high baseline GGT was an independent factor for OS in patients with metastatic renal cell carcinoma treated with nivolumab (P = 0.0345), and elevated GGT after treatment was an independent factor for progression‑free survival (PFS) (P = 0.0276) and OS (P = 0.0160).⁷¹ The present study further provides evidence that, among patients with HCC treated with PD‑1/PD‑L1 inhibitors, those with lower baseline GGT had longer PFS (P = 0.002), and that high baseline serum GGT is a poor prognostic factor for HCC patients. Currently, studies on the impact of elevated GGT levels in patients with cancer receiving immunotherapy are limited, and more evidence from diverse tumor types and larger sample sizes is needed.

ALP is an extracellular nucleotide enzyme widely present in human tissues and is composed of multiple isoenzymes. It hydrolyzes adenosine triphosphate into adenosine, adenosine monophosphate, and adenosine diphosphate extracellularly, and is involved in nucleotide metabolism. ALP is associated with inflammation, vascular calcification, metabolic syndrome, and cardiovascular diseases.⁷² ALP is a marker of pluripotent stem cells; therefore, ALP expression may promote the generation of cancer stem cells and increase tumor cell motility and migratory capacity.⁷³ Namikawa et al showed that high ALP (HR 1.808, P = 0.044) was an important independent poor prognostic factor for patients with advanced gastric cancer.⁷⁴ Compared with normal brain tissue, ALP exhibits higher activity in meningiomas, but not in gliomas, and higher ALP activity may serve as a surrogate marker for meningioma grading.⁷⁵ Patients with HCC who have elevated ALP and GGT levels have significantly lower OS and tumor-free survival than those with ALP and GGT within the normal range.⁷⁶ Our study demonstrated that patients with baseline ALP < 215 U/L had significantly longer PFS than those with ALP ≥ 215 U/L (HR = 0.246, P = 0.014, for ALP < 215 U/L vs ALP ≥ 215 U/L), indicating that elevated baseline ALP (≥ 215 U/L) is a risk factor for worse PFS in patients with HCC receiving PD‑1/PD‑L1 inhibitors.

While several recent studies have explored the prognostic value of peripheral blood markers in HCC patients receiving immunotherapy, our study offers several distinct contributions. First, we provide a comprehensive evaluation of 14 peripheral blood parameters in a Chinese cohort, with cutoff values determined using X‑tile software. Unlike previous studies that predominantly relied on composite scores such as NLR, PLR, or GLR, our analysis focused on individual markers. Notably, GGT demonstrated an exceptionally high HR of 8.100 (P = 0.002), suggesting that GGT may be a particularly powerful prognostic indicator. Second, we demonstrate that lung metastasis (HR = 2.823, P = 0.011) is an independent prognostic factor, highlighting the need for metastasis site‑specific stratification beyond the conventional binary classification of distant metastasis. Third, because all markers are derived from routine blood tests and clinical assessments, they are cost‑effective and readily applicable in routine clinical practice, particularly in resource‑limited settings.

Despite its clinical value, this study has several limitations. This study is a single-center retrospective analysis with a relatively small sample size, and all enrolled patients were Chinese, which may introduce selection bias. Furthermore, the number of objective response events was limited, with 2 CR and 7 PR cases, which substantially constrains the reliability of efficacy‑associated analyses and reduces the statistical power of the multivariable survival modeling. We acknowledge that the cutoff values for continuous variables were determined using X‑tile software within the same cohort used for survival analysis, without further validation. This approach, while commonly employed in exploratory prognostic studies, carries an inherent risk of overfitting and may inflate the statistical significance of the observed associations. The optimal cutoffs identified in our study should therefore be considered hypothesis‑generating rather than definitive, and require validation in independent cohorts. In this retrospective study, a major limitation is the high proportion of patients (82.2%) who received concomitant targeted therapy in combination with PD‑1/PD‑L1 inhibitors, which could have interfered with the evaluation of immunotherapy efficacy and is primarily attributable to the real‑world nature of the study design. Furthermore, patients received different PD‑1/PD‑L1 inhibitors and were treated across different lines of therapy, which introduced additional heterogeneity. This precludes a clear distinction between the prognostic effects attributable to immunotherapy alone versus those derived from the combination regimen. Therefore, the prognostic factors identified in this study should be interpreted as reflecting the outcomes associated with PD‑1/PD‑L1 inhibitor‑based therapy (including both monotherapy and combination regimens) rather than immunotherapy alone. Nevertheless, our findings inform the prognosis of patients receiving PD‑1/PD‑L1 inhibitors in the context of current real‑world practice, where combination therapy is the predominant treatment paradigm. In this retrospective cohort, baseline AFP and LDH values were missing for a considerable proportion of patients, precluding their inclusion in the multivariate analysis. Given that AFP is a well‑established prognostic biomarker in HCC, previously linked to tumor aggressiveness and poor survival, its omission from the predictive model constitutes a notable limitation. Future prospective studies with comprehensive data collection should incorporate AFP alongside the markers identified in our study to establish a more robust and clinically applicable prognostic model. In addition, several important prognostic factors were not fully incorporated in our analysis. Portal vein tumor thrombus and direct measures of tumor burden (eg, total tumor volume, sum of target lesion diameters) were not separately analyzed, as these data were not consistently documented across all patients. Prior TACE or other systemic therapies before PD‑1/PD‑L1 inhibitor initiation were also not systematically recorded in sufficient detail for meaningful subgroup analysis. Addressing these data gaps in future prospective studies would further strengthen the prognostic model. Due to the relatively short follow-up period, sufficient mortality events were not recorded; therefore, overall survival (OS) was not included as an endpoint. Nevertheless, PFS remains a clinically meaningful endpoint that reflects the time to tumor progression and provides valuable information on the early efficacy of immunotherapy. However, we acknowledge that OS is the gold‑standard endpoint for evaluating long‑term survival benefit. Extended follow‑up is ongoing, and we plan to report OS data in future updates. Furthermore, because of the presence of atypical response patterns, such as pseudoprogression during immunotherapy, the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 may underestimate the benefit of immunotherapy. Future multicenter prospective studies with larger sample sizes and longer follow-up are needed to provide more robust evidence. Blood components in the human body can directly or indirectly reflect physiological and pathological changes, and are therefore of great significance. Current techniques for blood sample analysis are well-established. Biomarkers obtained from clinical features and peripheral blood that are inexpensive and easily accessible may further optimize immunotherapy; however, blood sample analysis is susceptible to interference from multiple factors. Therefore, prognostic markers based on peripheral blood should be used in combination with other indicators to assess the prognosis. Despite these limitations, we believe our findings provide valuable preliminary evidence for identifying HCC patients most likely to benefit from PD‑1/PD‑L1 inhibitor‑based therapy. The markers we identified are derived from routine clinical practice, making them readily applicable in resource‑limited settings without additional cost.

Conclusion

In summary, the present study suggests that baseline BMI, Child‑Pugh class, line of therapy, distant metastasis, bone metastasis, ALP, and TBIL are factors associated with the response to PD‑1/PD‑L1 inhibitor therapy in patients with HCC. Lung metastasis, best response, baseline Hb, ANC, ALP, GGT, and CRP are independent prognostic factors for PFS in patients with HCC treated with PD‑1/PD‑L1 inhibitors, underlining the importance of incorporating these factors into prognostic assessment. To some extent, this study provides insights for clinicians to identify patients who are likely to respond to immunotherapy, and serves as a basis for future research on biomarkers in immunotherapy.

Funding Statement

This work was supported by the Zhejiang Provincial Natural Science Foundation of China (No. LY24H160004).

Abbreviations

Ref, reference; BMI, body mass index; CNLC, China Liver Cancer Staging; ECOG-PS, Eastern Cooperative Oncology Group Performance Status; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease; WBC, white blood cell count; RBC, red blood cell count; Hb, hemoglobin; PLT, platelet count; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; AMC, absolute monocyte count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALP, alkaline phosphatase; GGT, gamma-glutamyl transferase; CRP, C-reactive protein; ALB, albumin; TBIL, total bilirubin.

Data Sharing Statement

The datasets generated and/or analyzed during the present study are available from the corresponding author upon reasonable request.

Ethics Approval and Consent to Participate

The study was performed in accordance with the Declaration of Helsinki and relevant guidelines and regulations. This study was approved by the Institutional Review Board of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (approval number: 2023-0008). The requirement for informed consent was waived because this was a retrospective study using de-identified medical records, which posed minimal risk to participants. All data were anonymized before analysis to ensure patient confidentiality.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that there is no conflicts of interest in this work.