Comparative Analysis of the Immune Profiles of Women and Men With Hepatocellular Carcinoma Authors: Zuzana Macek Jilkova, Julien Ghelfi, Lucile Dumolard, Valerie Sinniger, Nour El Houda Mehdi, Christian Sengel, Yann Teyssier, Charlotte Costentin, Thomas Decaens Published: 2026 ABSTRACT ABSTRACTIntroductionHepatocellular carcinoma (HCC) is the most common primary liver cancer and a major global health concern, with a higher incidence in men than in women. In the era of immunotherapy, understanding sex‐based immunological differences is crucial.Methods/PatientsThis study investigated sex‐specific differences in anti‐tumor immunity in HCC through immune profiling of liver biopsies and blood samples. A prospective cohort of 101 patients was analyzed by multiparametric flow cytometry to assess lymphocyte populations and immune checkpoint molecule expression on tumoral and nontumoral biopsies and blood samples. Then, a retrospective cohort of 56 patients was analyzed for circulating immune markers and functional assessment.ResultsOur findings indicate that while overall intrahepatic immune composition was similar between sexes, female patients exhibited higher intra‐tumoral PD‐1high+CD8+ T cell frequency compared to men (13.3% ± 2.6% vs. 6.5% ± 0.7%, p = 0.0127), suggesting greater T cell exhaustion. Additionally, NK cells expressing CTLA‐4 and ICOS were more frequent in females. Circulating immune profiles also differed, with female patients showing increased CTLA‐4+ CD4+ and CD8+ T cells (p = 0.0005 and p = 0.0164) and reduced IFNγ expression, mainly in central memory CD8+ T cells.ConclusionsThese findings highlight the importance of considering sex‐related differences in the context of immunotherapy for HCC and underline the need for further investigation in larger cohorts. Introduction Hepatocellular carcinoma (HCC), the most common primary liver cancer, represents a major global health burden, with significant disparities in incidence and outcomes between men and women. Worldwide, men are disproportionately more affected by HCC, with a male‐to‐female incidence ratio ranging from 1.1 to 5, depending on the region [1, 2, 3], which contributes to the overall shorter life expectancy of men [4]. Gender and sex differences influencing HCC prevalence, development, and progression arise from a complex interplay of behavioral and biological factors, including a higher prevalence of HCC risk factors (such as viral hepatitis infections, intravenous drug use, and excessive alcohol consumption among men), as extensively reviewed in recent literature [2, 5]. One key aspect of HCC development and progression is the antitumor‐immune interaction, which may vary between sexes, as the strength of the immune response differs based on sex, shaping tumor evolution and response to therapy. Notably, estrogens have been shown to exert a protective effect against HCC development, while androgens promote tumor growth [6, 7, 8]. However, most HCC patients in Europe are diagnosed at an advanced age—typically around 70 years—when sex hormone levels are significantly reduced. This raises important questions about the persistence of sex‐related immune differences in older patients and their impact on disease progression and treatment efficacy. Immune‐related sex differences are particularly relevant in the era of immunotherapy, which has revolutionized the treatment landscape of HCC. Current first‐line systemic treatment for advanced HCC consists of immune checkpoint inhibitors (ICIs), targeting the programmed cell death protein 1/Programmed cell death‐ligand 1 (PD‐1/PD‐L1) pathway in combination with anti‐cytotoxic T‐lymphocyte associated protein 4 (CTLA‐4) and anti‐vascular endothelial growth factor (VEGF) agents [9]. Importantly, accumulating evidence suggests that sex differences influence immunotherapy responses across several cancer types [10, 11]. For instance, younger female patients often exhibit reduced responsiveness to PD‐1 blockade compared to their male counterparts, a phenomenon largely attributed to variations in the tumor immune microenvironment [12, 13]. Notably, a meta‐analysis revealed that men tend to benefit more from both CTLA‐4 inhibitors and PD‐1 inhibitors than women with melanoma or lung cancer. Moreover, the overall survival of patients treated with CTLA‐4 inhibitors appears to be even more influenced by sex differences than treatment with PD‐1 inhibitors [14]. Balcar et al. recently published a meta‐analysis examining sex‐related differences in the efficacy and safety of immunotherapy for HCC [15]. Pooled subgroup analyses showed a significant survival benefit in men, whereas in women, the survival benefit did not reach statistical significance. Notably, female patients only showed a significant benefit in the RATIONALE‐301 trial (PD‐1 inhibitor), while men demonstrated a survival advantage in three of the five studies (IMbrave150, KEYNOTE‐240, and LEAP‐002). However, the estimated treatment effect difference between sexes was 0.07 (95% CI 2.02–2.16), suggesting no significant sex‐based differences in treatment efficacy [15]. This underscores the need to better characterize the immune landscape of HCC, particularly in the context of sex‐based differences. The immune contexture of HCC before treatment has emerged as one of the most promising predictive markers for immunotherapy response [9, 16, 17], highlighting the importance of understanding the immunological distinctions between male and female patients. In general, women exhibit more robust cell‐mediated and humoral immune responses to antigenic challenges compared to men [18, 19]. While some immunological differences are present throughout life, others are more pronounced after puberty and decline after menopause (or andropause) [20]. These differences are particularly relevant in the context of lymphocyte‐mediated immunity, as lymphocytes play a pivotal role in anti‐tumor responses and immune surveillance. The expression of immune checkpoint molecules (ICMs), which regulate T cell activation and exhaustion, is influenced by sex‐based immunological factors [10]. Understanding how these mechanisms differ between men and women is essential for optimizing immunotherapeutic strategies for HCC. In this study, we perform phenotyping of immune cells from liver biopsies, focusing on both intra‐tumoral and nontumoral compartments, as well as circulating immune cells. By characterizing lymphocyte populations and their expression of ICMs, we aim to decipher the sex‐dependent variations in anti‐cancer immunity within HCC. Our results highlight the potential relevance of sex as a factor in ICI treatment outcomes. Patients and Methods Patients and Sample Processing A total of 101 patients diagnosed with HCC were prospectively enrolled in Cohort 1, prior to treatment at the Hepato‐Gastroenterology and Digestive Oncology Department, CHU Grenoble‐Alpes. Fresh blood samples were collected before core liver biopsy for multiparametric flow cytometry analysis. Liver biopsies, including tumor and nontumor tissues, were divided into two parts: one for histological assessment by experienced liver pathologists to confirm HCC diagnosis and determine biopsy location within or outside the tumor, and the other for phenotypic and functional immunological analyses. Analysis were processed within 4 h of collection to ensure optimal sample integrity. Cohort 2 was retrospective, comprising 56 patients diagnosed with early HCC. Peripheral blood mononuclear cells (PBMCs) were isolated from ~8 mL of blood samples using Ficoll density gradient centrifugation. The isolated PBMCs were subsequently cryopreserved in liquid nitrogen until analysis. This study was conducted in accordance with the Declaration of Helsinki and French legislation governing local sample collection (DC‐2014‐2295) and approved by the Ethics Committee of CHU Grenoble‐Alpes. Written informed consent was obtained from all participants. All experiments involved primary human samples; no cell lines were used. Flow Cytometry Analyses of Fresh Samples: Cohort 1 Immediately following liver biopsy, both tumor and nontumor tissue samples were transferred into Hypothermosol FRS solution and transported to the research laboratory. Biopsies were weighed, and cell suspensions were then obtained through mechanical disruption of the tissues. Fresh peripheral blood and intrahepatic cell suspension samples were then divided into two separate tubes for immunostaining. For surface marker analysis, immune cells were stained with a panel of anti‐human antibodies, including: CD45‐APC/Cy7 (clone HI30, BioLegend), CD3‐PerCP‐Cy5.5 (clone UCHT1, BioLegend), CD56‐BV605 (clone HCD56, BioLegend), CD16‐AF700 (clone 3G8, BioLegend), CD8‐PE/Cy7 (clone RPA‐T8, BD Biosciences), CD69‐PE (clone FN50, BioLegend), CTLA‐4‐BV421 (clone BNI3, BioLegend), PD‐1‐BV711 (clone EH12.2H7, BioLegend), LAG3‐PE (clone 3DS223H, eBioscience), ICOS‐BV785 (clone C398.4A, Biolegend), OX40‐FITC (clone Ber‐ACT35, BioLegend), 4‐1BB‐BV421 (clone 4B4‐1, BioLegend), TIM3‐BV785 (clone F39‐2E2, BioLegend). Dead cells were excluded using the PE‐TexasRed (LIVE/DEAD Fixable Red Dead Cell Stain, Invitrogen). To fix the cells, FluoroFix Buffer (BioLegend) was applied. Fluorescence minus one (FMO) controls were used to determine positive immune cell populations. Data acquisition was performed on a BD‐LSRII flow cytometer (BD Biosciences), using BD FACSDiva 6.3.1 software for collection. Flow cytometry data were subsequently analyzed using FCS Express 7 Flow software (De Novo Software), applying the gating strategy outlined in Figure S1A, and as previously established [21, 22]. The estimated number of immune cells per mg of tissue was calculated based on the original weight of each liver biopsy prior to processing. The mean weight was 7.4 mg for tumoral biopsies and 8.9 mg for nontumoral biopsies. The internal detection limit to exclude analysis was defined as samples with fewer than 200 CD45+ events. All samples in this cohort exceeded this threshold with an average of 8197 CD45+ lymphocytes per liver biopsy. Flow Cytometry Analyses of PBMCs: Cohort 2 PBMCs were thawed and seeded at a density of 0.3 × 106 cells per well in a 96‐well round‐bottom plate. Cells were left unstimulated or stimulated with a Phorbol 12‐myristate‐13‐acetate (PMA) and ionomycin (IONO) activation cocktail with Brefeldin A (BioLegend) for 5 h. PBMCs were first stained with a panel of antibodies targeting surface markers: CD45‐APC/Cy7 (HI30, BioLegend), CD3‐PerCP‐Cy5.5 (UCHT1, BioLegend), CD56‐BV605 (HCD56, BioLegend), CD19‐BUV805 (HIB19, BD Biosciences), CD8‐PE/Cy7 (SFCI21Thy2D3, BD Biosciences), CD4‐BV510 (SK3, BioLegend), CD69‐BV421 (FN50, BioLegend), CD45RA‐BUV395 (HI100, BD Biosciences), CCR7‐BUV661 (3D12, BD Biosciences), CXCR3‐BUV563 (1C6/CXCR3, BD Biosciences). For intracellular marker detection, cells were fixed, permeabilized, and stained with antibodies targeting intracellular cytokines: Interferon (IFN)γ‐PE (B27, BD Biosciences) and tumor necrosis factor (TNF)α‐Red718 (MAb11, BD Biosciences). Dead cells were excluded using the Zombie Green Fixable Viability Kit. FluoroFix Buffer (BioLegend) was used for fixation. Fluorescence minus one (FMO) controls were included to define positive immune cell populations. Samples were acquired using a FACSymphony flow cytometer, and data were collected with BD FACSDiva 6.3.1 software. Flow cytometry data were analyzed via supervised analysis using FCS Express 7 Flow software (De Novo Software). Cytokine Secretion by LUMINEX Technology To assess cytokine secretion from patients' PBMCs following in vitro stimulation, PBMCs were thawed and plated at a density of 0.3 × 106 cells per well in a 96‐well round‐bottom plate. Cells were either left unstimulated or activated with a Phorbol 12‐myristate‐13‐acetate (PMA) and ionomycin (IONO) cocktail (without Brefeldin A; BioLegend) for 5 h. Culture supernatants were collected and stored at −20°C until analysis. To evaluate the effector cell secretome, IFN‐γ and TNF‐α levels were quantified using LUMINEX technology. Assays were performed according to the manufacturer's instructions and measured with the MAGPIX 200 instrument using xPONENT software. Statistical Analysis Analyses were performed using the statistical software GraphPad Prism 10.4.1 (GraphPad Software). Normal distribution was tested by D'Agostino–Pearson omnibus normality test. When data from either cohort were not normally distributed, the Mann–Whitney test was performed to compare two groups (M vs. F group). p value of < 0.05 was considered significant. Data that follow a normal distribution are presented as the mean ± standard error (SE), whereas data that do not exhibit a normal distribution in either cohort are reported as medians with interquartile ranges (IQR). Results Among the 101 patients included in Cohort 1, 87 (86%) were men and 14 (13.9%) were women, Table 1. The median age was comparable between groups, with men having a median age of 69 years [IQR: 65–76] and women 70 years [IQR: 60–76]. The etiology of liver disease differed between sexes. Excessive alcohol consumption was more prevalent among male HCC patients than female patients (68.9% vs. 21.4%), whereas viral hepatitis was more common in women (50%) compared to men (12.6%). Biochemical parameters showed that male HCC patients had higher bilirubin levels than females (p = 0.0164). However, albumin levels, Child‐Pugh score, and transaminase levels (ALT and AST) were similar between sexes, suggesting a comparable degree of chronic liver disease. Alpha‐fetoprotein (AFP) level did not differ between men and women. TABLE 1Cohort 1: HCC patient characteristics.MalesFemales p n = 87 n = 14Age, y; median [IQR]69 [65–76]70 [60–76]0.7871BMI; median [IQR]27.1 [24.6–30.3]22.7 [16.4–29.2]0.0703Etiology of liver diseaseAlcohol; number (%)60 (68.9%)3 (21.4%)Metabolic; number (%)41 (47.1%)3 (21.4%)HCV/HBV; number (%)11 (12.6%)7 (50.0%)Unknown; number (%)0 (0%)1 (7.1%)Bilirubin, mg/dL; median [IQR]16.0 [10.0–24.0]11.2 [5.98–15.3]0.0164* Albumin level (g/dL); median [IQR]37.5 [32.9–41.3]36.9 (30.7–40.0]0.5412Child Pugh Score, A/B/C; number (%)74/12/1 (85/14/1%)12/2/0 (85/14/0%)> 0.9999ALT (U/L); median [IQR]40.0 [30.0–61.0]35.0 [18.5–60.3]0.3606AST (U/L); median [IQR]48.0 [35.0–71.5]34.5 [23.5–82.0]0.2034Alpha fetoprotein, ng/mL; median [IQR]13.1 [4.4–137]6.1 [3.7–450]0.5283Liver fibrosis, METAVIR score; F0‐F2; F3‐F4; number (%)19 (22%), 68 (78%)5 (36%), 9 (64%)0.0578Largest size of HCC, mm; median [IQR]33.0 [22.0–55.0]20.5 [14.8–32.3]0.0152* Number of HCC nodules/liver; median [IQR]1 [1–3]1 [1–2]0.1006Macrovascular invasion, Yes; number (%)17 (19.5%)2 (14.3%)0.7355Extrahepatic metastasis, Yes; number (%)13 (14.9%)0 (0%)0.2050BCLC stage, 0/A/B/C/D; number (%)7/36/19/20/5 (8/41/22/23/6%)5/5/1/3/0 (36/36/7/21/0%)0.0388* Previous Treatment, Yes; number (%)32 (36.8%)3 (21.4%)0.3687Surgical resection; number (%)4 (4.6%)0 (0%)Thermal ablation; number (%)4 (4.6%)0 (0%)Transarterial chemoembolization; number (%)18 (20.7%)1 (7.1%)TKIs; number (%)6 (6.9%)2 (14.3%) Note: Values are presented as median (IQR, 25th–75th percentile).Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; BCLC, Barcelona Clinic Liver Cancer system; TKIs, tyrosine kinase inhibitors.* p‐value < 0.05. Cohort 1: HCC patient characteristics. Note: Values are presented as median (IQR, 25th–75th percentile). Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; BCLC, Barcelona Clinic Liver Cancer system; TKIs, tyrosine kinase inhibitors. p‐value < 0.05. Tumor characteristics varied by sex, with male patients having significantly larger tumors (33 mm [IQR: 22–55] vs. 20.5 mm [IQR: 14.8–32.3], p = 0.0152). The majority of patients in both groups had monofocal HCC, no macrovascular invasion, and no extrahepatic metastases. However, according to the BCLC staging algorithm, male patients were classified into slightly more advanced HCC stages compared to female patients, Table 1. Sex‐Specific Variations in the Tumoral and Nontumoral Microenvironment Fresh liver biopsies (tumor and nontumor) were analyzed using multiparametric flow cytometry to characterize major lymphocyte populations and ICM expression. The frequencies of T cells, CD4+ T cells, CD8+ T cells, natural killer (NK) cells, and NKT cells within the CD45+ compartment were comparable between male and female HCC patients in both tumoral and nontumoral liver tissue (Figure 1A,B). Similarly, the estimated number of lymphocytes (normalized to original biopsy weight) of either tumor or nontumor tissue did not significantly differ between the two groups (Figure 1C,D). The proportion of activated lymphocytes, identified by CD69 expression, was also similar between male and female patients across both tumor and nontumor liver tissues (Figure 1E,F). Immunophenotyping of lymphocytes in tumoral and nontumoral liver tissue from male versus female HCC patients. Multiparametric flow cytometry was used to analyze the frequencies of key lymphocyte populations—T cells, CD4+ T cells, CD8+ T cells, natural killer (NK) cells, and NKT cells—within the CD45+ compartment, comparing male and female HCC patients. (A) Frequencies of immune cell populations gated from CD45+ lymphocytes in tumoral tissue. (B) Frequencies of immune cell populations gated from CD45+ lymphocytes in nontumoral tissue. (C) Estimated number of living immune cells per 1 mg of tumoral tissue. (D) Estimated number of living immune cells per 1 mg of nontumoral tissue. (E) Frequency of activated intra‐tumoral lymphocytes, identified by CD69 expression. (F) Frequency of activated lymphocytes in nontumoral tissue, identified by CD69 expression. (G) Frequencies of intra‐tumoral immune checkpoint molecule‐expressing cells among CD4+ T cells, CD8+ T cells, NK cells and NKT cells. (H) Frequencies of immune checkpoint molecule‐expressing cells in nontumoral tissue among CD4+ T cells, CD8+ T cells, NK cells, and NKT cells. Markers analyzed include PD‐1, PD‐1+high, LAG3, TIM3, CTLA‐4, ICOS, OX40, and 4–1BB. Statistical comparisons were performed using the Kruskal–Wallis test, followed by post hoc analysis of mean rank differences between male (M, n = 87) and female (F, n = 14) HCC patients, *p‐value < 0.05. Male patients are represented in green, and female patients in white. To further explore potential immune differences, we analyzed the frequency of CD4+ T cells, CD8+ T cells, NK cells, and NKT cells expressing various ICMs in fresh tumor and nontumor liver samples (Figure 1G,H). Intra‐tumoral CD8+ T cells expressing PD‐1high+ (Figure S1B) were significantly more prevalent in female patients than in males (13.3% ± 2.6% vs. 6.5% ± 0.7%, p = 0.0127) (Figure 1G). Similarly, intra‐tumoral T cells from female HCC patients generally exhibited a trend toward higher ICM expression compared to males (Figure 1G). In nontumoral liver tissue, NK cells expressing CTLA‐4 and inducible T cell costimulator (ICOS) were more frequent in female HCC patients than in males (Figure 1H, Figure S2A). The frequency of other ICM‐expressing cells in nontumoral liver tissue did not significantly differ between male and female patients. Together, these findings suggest sex‐specific variations in the intrahepatic immune microenvironment, with a higher frequency of ICM‐positive lymphocytes observed in female HCC patients. However, these differences remain modest, and the overall immune profile between male and female patients is largely comparable. Sex‐Specific Variations in the Circulating T Cells Next, we performed multiparametric flow cytometry on fresh blood samples to assess circulating lymphocyte populations. Consistent with findings in liver tissue, the frequencies of T cells, CD4+ T cells, CD8+ T cells, NK cells, and NKT cells within the CD45+ compartment were comparable between male and female HCC patients (Figure 2A). Likewise, the proportion of activated circulating lymphocytes expressing CD69 did not differ between the two groups (Figure 2B). Immunophenotyping of circulating lymphocytes from male versus female HCC patients. Multiparametric flow cytometry was used to assess the frequencies of key circulating lymphocyte populations and their immune checkpoint molecule expression in male and female HCC patients. (A) Frequencies of immune cell populations gated from CD45+ lymphocytes in blood. (B) Frequency of activated circulating lymphocytes (CD69+). (C) Frequencies of circulating immune checkpoint molecule‐expressing CD4+ T cells, CD8+ T cells, NK cells, and NKT cells. Statistical comparisons were performed using the Kruskal–Wallis test, followed by post hoc analysis of mean rank differences between male (M, n = 87) and female (F, n = 14) HCC patients, *p‐value < 0.05, ***p‐value < 0.001. Male patients are represented in green, and female patients in white. To further investigate potential immune differences, we analyzed the expression of ICMs on circulating CD4+ T cells, CD8+ T cells, NK cells, and NKT cells. Notably, female HCC patients exhibited a significantly higher frequency of CTLA‐4+ CD4+ T cells compared to males (2.00% ± 0.65% vs. 0.20% ± 0.03%, p = 0.0005) (Figure 2C, Figure S2B). Similarly, CTLA‐4+ CD8+ T cells were more prevalent in female patients than in males (p = 0.0164). However, ICM expression profiles in circulating NK and NKT cells were comparable between sexes. These findings suggest that female HCC patients have higher frequencies of CTLA‐4‐expressing T lymphocytes, which may influence their systemic immune response and potential sensitivity to anti‐CTLA‐4 immunotherapy. Sex‐Specific Variations in BCLC‐Comparable Male Versus Female HCC Patients To assess whether immune differences were influenced by higher tumor stage in male compared to female patients, we conducted a subcohort analysis, excluding male patients with advanced BCLC stage D (Figure S3, Table S1). Notably, no female patients in the cohort had stage D, which contributed to the difference in BCLC stage observed in the original Cohort 1. This adjustment allowed for a comparison between male and female patients with comparable BCLC stages (Table S1). Strikingly, the observed immune differences remained consistent and even became more pronounced (Figure S3A–L). PD‐1high+ CD8+ T cells were significantly more prevalent in females than males (p = 0.0112) (Figure S3B), reinforcing the notion of a sex‐specific intra‐tumoral immune landscape. Similarly, nontumoral NK cells expressing CTLA‐4 and ICOS remained more frequent in female patients (p = 0.0259, p = 0.0227) (Figure S3B). Moreover, analysis of circulating lymphocytes confirmed a significantly higher frequency of CTLA‐4+ CD4+ and CD8+ T cells in female patients compared to males (p = 0.0004, p = 0.0099) (Figure S3I–J). These findings suggest that sex‐based differences in ICM expression are independent of tumor stage and may reflect immune regulatory mechanisms that could influence response to immunotherapy in HCC patients. Sex‐Specific Variations in the Functionality of Circulating T Cells To further investigate sex‐specific differences in circulating T cells, we conducted an additional analysis of PBMCs from Cohort 2 (Table 2). This retrospective cohort included 56 patients, comprising 44 men (78.6%) and 12 women (21.4%). The median age was similar between the groups, with men having a median age of 72 years [IQR: 66–76] and women 73 years [IQR: 61–80]. Biochemical parameters and tumor characteristics were comparable between male and female patients. Most patients in both groups had monofocal HCC, no macrovascular invasion or extrahepatic metastases, and were classified as early stage (BCLC‐A) (Table 2). TABLE 2Cohort 2: HCC patient characteristics.MalesFemales p n = 44 n = 12Age, y; median [IQR]72 [66–76]73 [61–80]0.7483BMI; median [IQR]28.6 [25.9–32.6]28.3 [22.7–37.7]0.9668Etiology of liver diseaseAlcohol; number (%)27 (62.7%)10 (83.3%)Metabolic; number (%)22 (51.2%)4 (33.3%)HCV/HBV; number (%)11 (25.0%)6 (50.0%)Bilirubin, μmol/L; median [IQR]14.0 [9.2–18.7]9.0 [5.5–19.3]0.2290Albumin level (g/dL); median [IQR]40.0 [37.0–41.5]37.0 (34.9–39.8]0.0896Child Pugh Score, A/B; number (%)43/1 (97.7/2.3%)11/1 (91.6/8.3%)> 0.9999ALT (U/L); median [IQR]38.0 [27.0–56.0]24.0 [17.0–61.5]0.3586AST (U/L); median [IQR]42.0 [30.3–63.0]36.0 [28.5–92.5]0.8776Alpha fetoprotein, ng/mL; median [IQR]5.0 [3.0–11]5.6 [4.4–9.3]0.7085Liver fibrosis, METAVIR score; F0‐F2, F3, F4; number (%)14 (33%), 11 (25%), 19 (43%)4 (33%), 2 (17%), 6 (50%)0.9523Largest size of HCC, mm; median [IQR]20.0 [16.3–26.0)21.5 [18.0–29.8]0.4430Number of HCC nodules/liver, 1/> 1; number (%)40/4 (91/9%)12/0 (100/0%)0.5658Macrovascular invasion, Yes; number (%)0 (0%)0 (0%)> 0.9999Extrahepatic metastasis, Yes; number (%)0 (0%)0 (0%)> 0.9999BCLC stage, 0/A/B; number (%)15/28/1 (34/64/2%)4/8/0 (33/67/0%)> 0.9999Previous Treatment, Yes; number (%)20 (45.5%)2 (17%)0.1673Surgical resection; number (%)5 (11.4%)2 (17%)Thermal ablation; number (%)8 (18.2%)0 (0%)Transarterial chemoembolization; number (%)7 (15.9%)0 (0%)Very early recurrence, n (%)14 (31.8%)2 (16.7%)0.4751Death during follow‐up, n (%)11 (25.0%)4 (33.3%)0.7145 Note: Values are presented as median (IQR, 25th–75th percentile). Very early recurrence was defined as tumor occurring within 1 year post ablation.Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; BCLC, Barcelona Clinic Liver Cancer system. Cohort 2: HCC patient characteristics. Note: Values are presented as median (IQR, 25th–75th percentile). Very early recurrence was defined as tumor occurring within 1 year post ablation. Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; BCLC, Barcelona Clinic Liver Cancer system. Consistent with findings in Cohort 1, the frequencies of circulating T cells within the CD45+ compartment were comparable between male and female HCC patients (Figure S4A). To assess whether T cell differentiation varied by sex of HCC patients, we analyzed the distribution of differentiation stages. The subsets of TN: naïve, TCM: central memory, TEM: effector memory, and TEMRA: effector memory reexpressing CD45RA were distinguished by CD45RA and CCR7 expression in CD4+ and CD8+ T cells using flow cytometry (Figure 3A). Similarly, the cytokine expression and the capacity of T cells to respond to PMA/IONO stimulation was assessed (Figure 3B). Differentiation stages of circulating T cells from male versus female HCC patients and cytokine expression. Multiparametric flow cytometry was used to assess the frequencies and functionality of T cells in male and female HCC patients. (A) Gating strategy used for Cohort 2 analyses. CCR7+CD45RA+ TN: Naïve; CCR7+CD45RA− TCM: Central memory; CCR7−CD45RA−: Effector memory; CCR7−CD45RA+ TEMRA: Effector memory reexpressing CD45RA T cells. (B) Representative graphs of CD8+T cells gated from PBMCs that were either left unstimulated or stimulated for 5 h with a Phorbol 12‐myristate‐13‐acetate (PMA) and ionomycin (IONO) activation cocktail with Brefeldin A. (C) Frequencies of CD4+ T cells and CD8+ T cells according to differentiation stages. Statistical comparisons were performed using the Kruskal–Wallis test, followed by post hoc analysis of mean rank differences between male (M, n = 44) and female (F, n = 12) groups of HCC patients. (D, E) Frequencies of IFNγ‐expressing CD4+ T cells (D) and CD8+ T cells (E). (F) PBMCs were either left unstimulated or activated with PMA/IONO without Brefeldin A for 5 h. Culture supernatants were quantified to evaluate the levels of IFN‐γ. (G) Frequencies of IFNγ‐expressing CD8+ T cells according to differentiation stages: TN, TCM, TEM, TEMRA. Statistical comparisons were performed using the Wilcoxon pair test for nonstimulated vs. stimulated conditions: #p‐value < 0.05, ##p‐value < 0.01, ###p‐value < 0.001, ####p‐value < 0.0001. The Mann–Whitney test was used to compare male (M, n = 44) and female patients (F, n = 12): *p‐value < 0.05, **p‐value < 0.01. Male patients are represented in green, and female patients in white. The frequencies of CD4+ and CD8+ T cells at each differentiation stage were similar between male and female HCC patients (Figure 3C). However, in the absence of stimulation, IFNγ expression was slightly higher in CD4+ T cells from male HCC patients compared to females (Figure 3D). A significant increase was observed in IFNγ‐expressing CD8+ T cells in males compared to females (27.6 [IQR: 6.4–42.8] vs. 7.5 [IQR: 3.0–25.0], p = 0.0414) (Figure 3E), which was further confirmed by elevated IFNγ secretion in the male compared to the female group under nonstimulated conditions (18.8 [IQR: 9.8–29.4] vs. 7.3 [IQR: 3.5–12.0], p = 0.0019) (Figure 3F). No sex‐related differences were detected in the frequency of TNFα‐expressing T cells (Figure S4B) or in TNFα secretion (Figure S4C). Upon PMA/IONO stimulation, both male and female exhibited similar frequencies of T cells expressing IFNγ and TNFα, as well as similar cytokine secretion levels, indicating preserved functional responsiveness across sexes (Figure 3D–F, Figure S4B,C). To further investigate the sex differences in IFNγ expression within CD8+ T cells, we analyzed IFNγ levels across different CD8+ T cell differentiation stages (Figure 3G). Notably, IFNγ‐expressing CD8+ TCM cells were significantly more frequent in male HCC patients compared to females under nonstimulated conditions (24.4 [IQR: 5.1–41.8] vs. 3.6 [IQR: 1.7–19.8], p = 0.0048). Similarly, significant differences were observed in IFNγ‐expressing CD8+ TEMRA subsets (p = 0.0480). Furthermore, upon PMA/IONO stimulation, TN and TCM subsets exhibited increased IFNγ expression exclusively in male HCC patients. Overall, these findings highlight sex‐based differences in IFNγ expression within CD8+ T cell subsets, suggesting a sex‐specific circulating immune profile of HCC patients. Discussion In this study, we observed that female HCC patients exhibited a higher frequency of ICM‐positive intra‐tumoral T cells, with PD‐1high+CD8+ T cells being significantly more prevalent in females than males. Additionally, NK cells expressing CTLA‐4 and ICOS were more frequent in female patients. These findings suggest that female HCC patients may have a more immunosuppressed tumor microenvironment, characterized by increased ICM expression. Higher PD‐1 expression in intra‐tumoral CD8+ T cells could indicate a greater degree of T cell exhaustion, which may impact antitumor immunity and responsiveness to ICIs. In addition, analysis of circulating lymphocytes revealed that female patients exhibited a significantly higher frequency of CTLA‐4+ CD4+ and CD8+ T cells compared to males. This suggests that systemic immune suppression via CTLA‐4 may be more pronounced in female HCC patients, potentially influencing their sensitivity to CTLA‐4‐targeted therapies. Notably, a recent study found that nonresponders to anti‐PD‐L1/anti‐VEGF treatment in advanced HCC exhibited a higher percentage of CTLA‐4+ circulating lymphocytes compared to responders [23]; however, sex‐specific analyses were not conducted. Moreover, in our study, male patients demonstrated significantly higher IFNγ expression in circulating CD8+ T cells, particularly in the CD8+ TCM subset: long‐lived T cells that provide central immunosurveillance and play a critical role in antitumor immunity [24]. These differences suggest that male patients may have a more pro‐inflammatory circulating immune profile, which could impact tumor progression and response to therapy. On the other hand, reduced IFNγ production in CD8+ T cells is associated with T cell exhaustion [25], aligning with the increased ICM expression observed in female patients. Notably, Conforti et al. previously reported that the tumor microenvironment of women with advanced nonsmall cell lung cancer (NSCLC) was characterized by significantly greater T cell dysfunction status and higher expression of inhibitory ICM [13]. Similarly, tumors from female patients with nonmuscle‐invasive bladder cancer exhibited significantly elevated expression of immune checkpoint genes (CTLA4, PDCD1, LAG3, and ICOS) compared to male patients [26]. Consistent with these findings across other cancers, our results underscore the importance of considering sex as a biological variable in HCC research and treatment. While the observed differences in ICM expression and cytokine production are modest, they suggest that men and women may exhibit differential responses to immunotherapy. Importantly, it has been shown that previous HCC treatment can increase ICM expression and induce a more exhausted tumor microenvironment [27, 28]. In our cohorts, female patients—who displayed higher ICM expression and signs of exhaustion compared to male patients—were less likely to have received prior treatment. Therefore, the observed sex‐based immune differences may, in fact, be even more pronounced in treatment‐naïve datasets. Our study has certain limitations. First, the immune landscape of HCC may vary significantly based on the underlying liver disease etiology [29]. Our single‐institution cohorts are insufficient to fully investigate the associations with the background etiology of liver diseases, but they provide a strong rationale for future large‐scale, multi‐institutional studies to validate the observed sex‐based immune differences and their clinical significance. Second, in our cohorts, women represented the minority of HCC cases, which reflects real‐world epidemiology in European countries [1, 2]. However, this sex imbalance may limit the statistical power of our analyses, and more balanced cohorts would be preferable for exploring sex‐specific differences with greater precision. Furthermore, patients in our study presented with heterogeneous tumor stages and consequently received diverse therapeutic interventions. As a result, our study design does not allow us to directly assess the relationship between immune cell profiles and response to therapy, particularly immunotherapy, which remains a critical question for future research. Additionally, a more in‐depth evaluation of immune cell functional states within the tumor microenvironment is needed. Further research should explore the mechanistic drivers of these disparities and clinical trials should incorporate sex‐stratified analyses to refine immunotherapeutic strategies and optimize treatment outcomes for both male and female HCC patients. Conclusions To conclude, this study highlights sex‐related differences in the immune landscape of HCC patients. While the overall immune profile was largely comparable between sexes, female patients exhibited a higher frequency of ICM‐positive intra‐tumoral T cells and circulating CTLA‐4+ T cells, whereas male patients demonstrated greater IFNγ expression in circulating CD8+ T cells. These results suggest that sex‐based immune variations may influence HCC immunotherapy outcomes. Recognizing these differences is essential for optimizing personalized treatments and warrants further investigation. Author Contributions Zuzana Macek Jilkova: investigation, writing – original draft, funding acquisition, writing – review and editing, visualization, validation, formal analysis, project administration, supervision, resources, data curation, conceptualization. Julien Ghelfi: conceptualization, investigation, funding acquisition, writing – review and editing, supervision, resources, data curation, project administration, formal analysis, validation. Lucile Dumolard: data curation, formal analysis, validation, investigation, writing – review and editing. Valerie Sinniger: data curation, formal analysis, investigation, validation. Nour El Houda Mehdi: investigation, formal analysis, data curation. Christian Sengel: resources, project administration. Yann Teyssier: resources, project administration. Charlotte Costentin: conceptualization, investigation, validation, writing – review and editing, project administration, resources, supervision. Thomas Decaens: project administration, writing – review and editing, conceptualization, investigation, resources, supervision. Funding This research was funded by “Association Française pour l'Étude du Foie,” “Fonds Agir pour les Maladies Chroniques” and “La Ligue contre le cancer AURA”. Ethics Statement This study was conducted in accordance with the Declaration of Helsinki and French legislation governing local sample collection (DC‐2014‐2295) and approved by the Ethics Committee of CHU Grenoble‐Alpes. Written informed consent was obtained from all participants. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Table S1: Sub‐Cohort 1 (HCC patient characteristics (BCLC matched)). Figure S1: Multiparametric flow cytometry to phenotype immune cells in HCC patients. Figure S2: Sex‐specific differences in immune checkpoint molecule expression in male versus female HCC patients. Figure S3: Sex‐specific differences in immune checkpoint molecule expression in male versus female HCC patients, matched on BCLC stage. Figure S4: Analyses of circulating lymphocytes in HCC patients.