Skip to main content
Journal of Hepatocellular Carcinoma logoLink to Journal of Hepatocellular Carcinoma
. 2026 Nov 5;13:651068. doi: 10.2147/JHC.S651068

Immune Dynamics of Pathological Complete Response to Neoadjuvant Immunotherapy in Hepatocellular Carcinoma: Mechanisms, Biomarkers, and Prognostic Implications

Yunzhi Ling 1, Guoming Li 1, Yijun Liu 1, Weijun Li 1, Yantao Fu 1,✉
PMCID: PMC13746913  PMID: 43005931

Abstract

Hepatocellular carcinoma (HCC) is one of the most common malignancies worldwide, with 5-year postoperative recurrence rates as high as 70%, severely limiting long-term survival after surgical resection. The success of immune checkpoint inhibitors in advanced HCC has driven their expansion into the neoadjuvant setting, where pathological complete response (pCR) has emerged as a potential endpoint associated with improved outcomes, although its validation as a surrogate for overall survival requires further prospective evidence. This review systematically synthesizes current evidence on the immune dynamics underlying pCR following neoadjuvant immunotherapy in HCC. Pooled analyses report pCR rates of approximately 18–24%, while certain intensive regimens, such as dual immune checkpoint inhibitor combinations, have achieved rates up to 41% in selected populations, albeit with increased toxicity. The peripheral infiltration and expansion of high-frequency intratumoral T-cell clonotypes are associated with pathological response, although a causal relationship remains to be established. Baseline tumor-infiltrating lymphocyte density, tertiary lymphoid structures, and dynamic changes in circulating tumor DNA are emerging as promising predictive biomarkers. In pan-tumor analyses, patients achieving pCR demonstrate substantially improved relapse-free survival, with 3-year event-free survival rates ranging from 78.6% to 100%; however, HCC-specific prospective data remain limited. However, heterogeneity in pCR definitions, lack of standardized pathological assessment, and limited long-term follow-up data constrain the definitive qualification of pCR as a surrogate endpoint. Future directions include standardization of pCR assessment, multi-omics integration for improved prediction, biomarker-guided patient selection, and large-scale prospective validation. This review provides a comprehensive framework for understanding the immune mechanisms of pCR and its clinical implications, offering guidance for clinicians and researchers in the evolving landscape of neoadjuvant immunotherapy for HCC.

Keywords: hepatocellular carcinoma, neoadjuvant immunotherapy, pathological complete response, tumor immune microenvironment, biomarkers, prognosis

Introduction

Hepatocellular carcinoma (HCC) accounts for over 80% of all primary liver cancers and remains the fourth leading cause of cancer death worldwide.1 Surgical resection serves as the mainstay curative treatment for early-stage HCC, offering 5-year survival rates of approximately 75% in early-stage disease.2 However, postoperative recurrence remains a formidable challenge, with rates as high as 70% within 5 years of surgery, driven by the presence of microvascular invasion and undetectable micrometastases at the time of resection.2 Traditional adjuvant therapies have yielded limited success, underscoring the urgent need for more effective perioperative strategies.3

The remarkable success of immune checkpoint inhibitors (ICIs) in advanced HCC has catalyzed their investigation in the neoadjuvant setting.3 Neoadjuvant immunotherapy offers several theoretical advantages: the presence of an intact primary tumor provides a broader spectrum of tumor antigens, facilitating T-cell activation and expansion within the tumor microenvironment; it can downstage locally advanced tumors, converting unresectable disease into resectable disease; and it may eliminate micrometastatic disease, thereby reducing postoperative recurrence.2

The rationale for neoadjuvant immunotherapy in HCC rests on both biological and clinical foundations. Biologically, the presence of an intact primary tumor provides a broad repertoire of tumor antigens, which can prime and expand tumor-reactive T cells before micrometastatic niches become established. Neoadjuvant ICIs may thus induce durable immune memory and eliminate micrometastatic disease, thereby reducing postoperative recurrence. Clinically, early-phase trials have demonstrated that neoadjuvant immunotherapy can achieve pathological responses, including pCR, and that pCR is associated with improved relapse-free survival. However, the existing evidence base remains limited: most studies are single-arm Phase I/II trials with small sample sizes, heterogeneous patient selection criteria, variable treatment regimens, and inconsistent definitions of pathological response. Few trials include control groups or long-term overall survival data. Therefore, while the neoadjuvant approach is biologically compelling, its clinical benefit requires validation in larger, prospective randomized trials with standardized endpoints. A critical synthesis of the current evidence is needed to identify gaps and guide future research.

Pathological complete response (pCR), defined as 100% tumor necrosis in the resected surgical specimen, represents the most profound depth of tumor regression achievable with neoadjuvant therapy.4 In HCC, pCR has emerged as a potential endpoint associated with improved long-term outcomes, but it has not yet been validated as a surrogate for overall survival in prospective trials. Landmark cross-trial analyses have demonstrated that pCR following neoadjuvant immunotherapy is associated with significantly prolonged relapse-free survival.4 A 2025 meta-analysis of 14 clinical trials involving 252 patients further supported the potential clinical relevance of pCR as an efficacy endpoint, though definitive prospective validation is still needed.5 These pooled estimates reflect heterogeneous patient populations and treatment regimens and should not be directly compared with rates from individual trials. However, the immune mechanisms associated with pCR are not fully understood, and causality has not been definitively established. Why do some patients achieve complete tumor eradication while others derive minimal benefit? This question centers on the complex dynamics of the tumor immune microenvironment and its remodeling during neoadjuvant immunotherapy.6

This review provides a systematic synthesis of current evidence on the immune dynamics of pCR following neoadjuvant immunotherapy in HCC. We examine the immunological mechanisms underlying pCR, explore emerging biomarkers for predicting pCR, evaluate the prognostic significance of pCR, and discuss current challenges and future directions for optimizing neoadjuvant immunotherapy in HCC.

Search Strategy and Selection Criteria

We conducted a comprehensive literature search in PubMed, Embase, Web of Science, and the Cochrane Library from inception to December 2026. The search terms included “hepatocellular carcinoma”, “neoadjuvant immunotherapy”, “pathological complete response”, “immune dynamics”, “biomarkers”, and “prognosis”, combined with Boolean operators. Inclusion criteria were: (1) clinical studies, reviews, or meta-analyses; (2) studies reporting pCR or immune-related outcomes after neoadjuvant immunotherapy in HCC; (3) English-language publications. Exclusion criteria were: (1) case reports, conference abstracts, or editorials; (2) studies not focused on HCC or pCR; (3) non-English articles. After removing duplicates, titles and abstracts were screened, followed by full-text review. A total of XX studies were finally included.

Pathological Complete Response in Neoadjuvant Immunotherapy for HCC

Definition and Assessment of pCR

Pathological response is measured as the percentage of non-viable tumor in the resected surgical specimen. Pathological complete response (pCR) refers to 100% tumor regression, indicating the complete absence of viable tumor cells in the resected specimen.4 For context, major pathological response (MPR) is conventionally defined as at least 70% tumor regression, although some studies use a ≥90% threshold; however, this review focuses on pCR as the primary endpoint.7 Despite the uniform definition of pCR as 100% tumor regression, several methodological factors introduce substantial heterogeneity in its assessment across studies. First, specimen sampling may vary: the number of sections examined, the sampling of the tumor bed, and the inclusion of peritumoral tissue can all affect the detection of viable tumor cells. Second, pathological assessment is subject to inter-observer variability, particularly in distinguishing treatment-related necrosis from viable tumor, and in interpreting immune-related pathological features. Third, the use of central pathology review vs local institutional review can lead to different pCR rates; central review is generally more standardized but is not universally implemented. These differences directly affect cross-study comparisons and may partially explain the wide range of pCR rates reported in the literature. Standardized protocols for specimen handling, sectioning, and central pathology review are urgently needed to improve the comparability of pCR data across trials.

Incidence of pCR Across Different Neoadjuvant Immunotherapy Regimens

The reported incidence of pCR varies substantially across studies, but this variation must be interpreted in light of differences in study design, patient selection, and treatment regimen. The main sources of data are summarized below.

Pooled analyses and meta-analyses: A 2024 meta-analysis of 255 patients from 11 studies reported an overall MPR rate of 47% (95% CI 31–70%) and a pCR rate of 22% (95% CI 14–36%).8 A separate cross-trial, patient-level analysis of 111 patients from five Phase I/II trials reported MPR in 32% and pCR in 18%.4 Another meta-analysis demonstrated that ICIs significantly increase the pCR rate (OR = 0.23, 95% CI 0.14–0.37, p < 0.00001) and MPR rate (OR = 0.47, 95% CI 0.32–0.70, p = 0.0002).9 Consistent with these findings, a recent meta-analysis by Zhao et al reported a pooled pCR rate of 20% (95% CI 14–30%) for ICI–TKI combinations,10 and Yuza et al summarized that neoadjuvant ICI-based regimens achieve MPR rates of 20–42%.11 These pooled estimates should be interpreted with caution because they combine different treatment regimens and patient populations.

Regimen-specific data:

  • Single-agent ICIs: pCR rates range from 12.9% to 19% in patients with resectable HCC.10

  • ICI–TKI combinations: pCR rates of approximately 20% have been reported, often in patients with more advanced disease.11

  • Dual ICI combinations (eg, nivolumab plus ipilimumab): The PRIME-HCC study reported a pCR rate of 22%. In a separate cross-trial analysis, dual ICI therapy achieved a pCR rate of 41% (95% CI 29–59%) and an MPR rate of 42% (95% CI 26–67), but with severe treatment-related adverse events in 41% of patients (95% CI 29–60%).12

  • Conversion therapy: pCR rates in this setting are reported separately (see pCR in Special Populations) and apply to patients with initially unresectable disease, a population distinct from those receiving upfront neoadjuvant therapy.

Therefore, the widely cited range of 18–24% for pCR reflects pooled data from heterogeneous studies, while higher rates such as 41% come from specific regimens in selected populations. Direct comparisons across these numbers should be avoided without adjusting for study design and patient selection. Table 1 summarizes the key clinical trials evaluating neoadjuvant immunotherapy in HCC. Notably, the trials summarized in Table 1 differ substantially in patient selection, treatment regimens, and pathological assessment, which limits direct comparisons and underscores the need for standardized prospective studies.

Table 1.

Summary of Major Clinical Trials Evaluating Neoadjuvant Immunotherapy in Hepatocellular Carcinoma

Study/Trial Regimen Sample Size pCR Rate MPR Rate Key Findings
D’Alessio et al (2024)4 Pooled analysis: anti-PD-1/PD-L1 ± anti-CTLA-4 111 18% 32% Median RFS 43.6 months; MPR associated with HR 0.27 for relapse (p=0.003)
Tian et al (2024)8 Meta-analysis: ICIs ± TKIs 255 22% (95% CI 14–36%) 47% (95% CI 31–70%) Pooled analysis of 11 studies
Tuersun et al (2024)9 Meta-analysis: ICIs ± TKIs 252 OR 0.23 (p<0.00001) OR 0.47 (p=0.0002) 14 trials; ICIs significantly increased pCR and MPR rates
Zhao et al (2023)10 Meta-analysis: anti-PD-1/PD-L1 276 13% (95% CI 8–18%) 30% (95% CI 22–38%) Single-agent ICI therapy
PRIME-HCC (Pinato et al, 2021)11 Neoadjuvant ipilimumab + nivolumab 35 22% 33% Phase Ib study; preliminary results suggest dual ICI feasibility
Marron et al (2022)13 Neoadjuvant cemiplimab (anti-PD-1) 21 – 20%* Phase II trial; significant tumor necrosis (>70%) in 20% of patients
CARES-009 (Wang et al, 2025)14 Perioperative camrelizumab + rivoceranib vs surgery alone 294 (148 vs 146) 3.4% vs NR 35.1% vs 7.5% (p<0.001) Randomized Phase 2/3; median EFS 42.1 vs 19.4 mo (HR 0.59)
Lin et al (2026)15 Nivolumab + ipilimumab 43 (24 surgical) NR 33.3% (>90% necrosis) 4-yr PFS 44%; 4-yr OS 60%
Dai et al (2026)16 FOLFOX-HAIC + tislelizumab 26 (22 surgical) 27.3% 45.5% ORR 57.7%; DCR 96.2%
Song et al (2026)17 Perioperative tislelizumab monotherapy 36 (33 surgical) 15.2% 33.3% Grade ≥3 AEs 11.1%; no surgical delays

Notes: The heterogeneity in MPR thresholds across studies (≥70% vs ≥90% tumor regression) and differences in trial designs should be considered when interpreting pooled estimates. Data are presented as reported in the original publications. *In the Marron et al study, significant tumor necrosis was defined as >70%, which corresponds to the MPR threshold used in other studies. Literature search updated to September 2026. NR, not reported; EFS, event-free survival; AEs, adverse events.

Abbreviations: pCR, pathological complete response; MPR, major pathological response; ICI, immune checkpoint inhibitor; TKI, tyrosine kinase inhibitor; RFS, recurrence-free survival; HR, hazard ratio; OR, odds ratio; CI, confidence interval.

Recent evidence highlights that the efficacy of ICIs in HCC is strongly constrained by poor immune infiltration and an immunosuppressive tumor microenvironment (TME), supporting rational combinations with other therapeutic modalities.18 For example, anti-angiogenic agents can normalize tumor vasculature and promote T-cell infiltration, while TKIs may modulate immune cell function and reduce immunosuppressive populations. Locoregional therapies such as radiotherapy or ablation can induce immunogenic cell death and enhance antigen presentation, thereby priming the TME for ICI response. These combination strategies may convert immune-excluded or immunosuppressive tumors into immune-inflamed tumors, facilitating T-cell clonal expansion, tertiary lymphoid structure formation, and ultimately increasing pCR rates. Therefore, rational combination design should be guided by baseline immune contexture and on-treatment immune remodeling, with the goal of maximizing pCR while minimizing toxicity.

More recent trials have further expanded the evidence base. The randomized Phase 2/3 CARES-009 trial demonstrated that perioperative camrelizumab plus rivoceranib significantly improved median EFS compared with surgery alone (42.1 vs 19.4 months; HR 0.59, 95% CI 0.41–0.85; p=0.0040) and achieved higher MPR rates (35.1% vs 7.5%; p<0.001), with a pCR rate of 3.4%.14 A single-arm Phase II trial of nivolumab plus ipilimumab in 43 patients with potentially resectable HCC reported major pathological response (>90% necrosis) in 8 of 24 surgical patients, with estimated 4-year PFS of 44% and OS of 60%.15 A single-arm Phase II trial of FOLFOX-HAIC plus tislelizumab in patients beyond the Milan criteria reported an MPR rate of 45.5% and a pCR rate of 27.3% among surgical patients.16 Finally, a Phase II trial of perioperative tislelizumab monotherapy in 36 patients reported an MPR rate of 33.3% and a pCR rate of 15.2%, with grade ≥3 adverse events in 11.1% and no surgical delays.17

pCR as a Surrogate Endpoint: Evidence and Controversies

The qualification of pathological response as a surrogate endpoint for survival is a critical question. While pCR is consistently associated with improved RFS, it is important to note that no prospective randomized trial has yet demonstrated that achieving pCR directly translates into improved overall survival. Therefore, pCR should currently be regarded as a promising but not yet validated surrogate endpoint. A landmark study by D’Alessio et al evaluated 111 patients treated with neoadjuvant ICIs across 12 tertiary referral centers in the United States, Europe, and Taiwan.4 After a median follow-up of 27.2 months, median RFS was significantly improved in patients with MPR (not reached vs 28.3 months; HR 0.27, 95% CI 0.11–0.69, p = 0.003) and with pCR (not reached vs 32.8 months).4 This study provided the first qualification of MPR as a putative surrogate endpoint for RFS following neoadjuvant immunotherapy for HCC. As detailed in Incidence of pCR Across Different Neoadjuvant Immunotherapy Regimens, a meta-analysis demonstrated that ICIs significantly increase both pCR and MPR rates,9 supporting the clinical relevance of these pathological endpoints. An accompanying editorial highlighted that the identification of 90% tumor regression as an optimal threshold could help standardize response assessment in future trials, while also noting important methodological considerations including the absence of a control group and the heterogeneity of treatment regimens across the pooled cohort.7

However, several controversies remain. First, the correlation between radiological response and pathological response is imperfect (R2 = 0.43, p < 0.0001), indicating that pathological assessment provides unique prognostic information not captured by imaging.4,19 Second, the durability of the surrogate relationship across different neoadjuvant regimens and patient populations needs confirmation in larger prospective studies. Furthermore, the lack of standardized pathological assessment protocols across centers and the absence of central pathology review in many studies remain significant barriers to the widespread adoption of pCR as a definitive surrogate endpoint.20 As noted in Definition and Assessment of pCR, the MPR threshold remains debated; however, this review focuses on pCR.

Immune Dynamics of pCR: Mechanisms and Microenvironment Remodeling

Baseline Immune Landscape: Predictors of pCR

The baseline tumor immune microenvironment plays a critical role in determining response to neoadjuvant immunotherapy. Patients who achieve pCR or MPR exhibit distinct immune profiles prior to treatment initiation.20

Tumor-infiltrating lymphocytes (TILs) represent one of the most consistent baseline predictive biomarkers.20,21 Higher densities of CD8+ T cells in pretreatment tumor biopsies are associated with improved pathological responses.21,22 Studies have demonstrated that tumors with higher baseline T-cell infiltration are more likely to undergo immune-mediated regression following ICI therapy. For patients with HCC scheduled to receive immunotherapy, TILs have been identified as potential predictive biomarkers of treatment response.20 Furthermore, the peripheral infiltration and expansion of high-frequency intratumoral T-cell clonotypes are critical drivers of pathological response to neoadjuvant PD-1 blockade in HCC.6

Tertiary lymphoid structures (TLS) have emerged as a particularly important predictive biomarker.21 TLS are ectopic lymphoid aggregates that form within or adjacent to tumors and serve as sites of local antigen presentation, lymphocyte activation, antibody production, and effector output. For patients with HCC scheduled to receive immunotherapy, intratumoral TLS has also been identified as a potential predictive biomarker of treatment response.20 In HCC treated with neoadjuvant immunotherapy, high intratumoral TLS density at the time of surgery is associated with pathologic response and improved relapse-free survival.23 TLS have been proposed to orchestrate a complete “intratumoral immune cycle”, enabling local antigen presentation, lymphocyte activation, antibody production, and effector output.23 Immunotherapy-induced TLS formation is associated with enhanced antitumor immunity.24

The liver’s unique immune environment presents both opportunities and challenges. The liver is characterized by a tolerogenic microenvironment that may dampen antitumor immune responses. The tumor immune microenvironment and the mechanisms of action of ICIs in early-stage HCC have been comprehensively described in recent reviews.3 However, the presence of the primary tumor during neoadjuvant treatment may overcome this tolerance by providing a broader spectrum of tumor antigens and promoting T-cell activation within the cancer microenvironment. The temporal progression of immune dynamics from baseline tumor microenvironment to pCR is summarized in Figure 1.

Figure 1.

Five-step HCC timeline from baseline immune microenvironment to pathological complete response. Flowchart sequence: Baseline HCC immune microenvironment, Neoadjuvant ICI therapy, Clonal expansion and infiltration, Microenvironment remodeling, Pathological complete response. Baseline HCC immune microenvironment shows a central cluster labeled HCC cell. Around it are immune cells labeled CD8 superscript plus T cell, Kupffer cell and T reg, plus a structure labeled TLS. Multiple labels PD-1 and PD-L1 are placed near immune cells around the tumor cluster. Neoadjuvant ICI therapy contains the text PD-1 slash PD-L1 blockade, with an antibody-like icon next to the label PD-1 and a downward arrow pointing to two immune cells shown interacting. Clonal expansion and infiltration shows multiple immune cells with curved arrows around them labeled Clonal expansion. A band at the bottom is labeled Peripheral blood and contains many small circular cells. An arrow labeled Infiltration points upward from Peripheral blood toward the immune cells. Microenvironment remodeling contains the text Cold right arrow Hot. A tumor cluster is surrounded by many immune cells and labeled elements including CAF, TLS and Macrophage. Pathological complete response shows a central area labeled pCR with surrounding text Necrosis slash Debris and No viable tumor cells, plus scattered immune cells around the central region.

Integrated mechanistic framework linking baseline immune status, treatment-induced immune remodeling, tumor regression, pCR, and recurrence/survival in HCC. The five-step timeline depicts: (1) baseline immune contexture; (2) ICI-induced T-cell activation and clonal expansion; (3) peripheral infiltration and intratumoral T-cell expansion; (4) remodeling from “cold” to “hot” tumor microenvironment; (5) achievement of pCR. Post-pCR recurrence or survival depends on sustained immune surveillance and absence of immune-evasive clones.

It is important to distinguish between three related but distinct concepts: (1) predictive biomarkers, which are baseline features associated with pCR; (2) mechanisms, which are treatment-induced immune changes that may causally contribute to pCR; and (3) consequences, which are outcomes that follow pCR, such as improved relapse-free survival. Most current evidence for immune features such as TILs, TLS, CAFs, and macrophages is correlative, and their causal role in driving pCR requires further functional validation.

Dynamic Changes in the Tumor Immune Microenvironment During Neoadjuvant Immunotherapy

Neoadjuvant immunotherapy induces profound and dynamic changes in the tumor immune microenvironment. These changes can be tracked through sequential biopsies or through paired pre- and post-treatment surgical specimens.

T-cell dynamics represent the most well-characterized immune changes associated with treatment response.6,23 Zhang et al analyzed longitudinal matched tumor and blood samples from 19 treatment-naïve HCC patients receiving neoadjuvant nivolumab, integrating T-cell receptor repertoire analysis with transcriptomic profiling.6 However, the small sample size (n = 19), single-arm design, and lack of independent validation in larger prospective cohorts limit the generalizability of these findings. Nonetheless, collectively, these studies support a model in which effective antitumor immunity depends on the expansion and trafficking of high-frequency intratumoral T-cell clonotypes, a process associated with pathological response. They demonstrated that elevated clonality of intratumoral TCR clonotypes post-treatment correlated with increased pathological tumor necrosis, predominantly driven by the most abundant top 1% of intratumoral TCR clonotypes.6 These dominant clonotypes demonstrated increased spatial sharing across tumor, non-tumorous liver, and peripheral blood compartments in nivolumab responders, and shared clonotypes were positively associated with tumor necrosis extent and cytolytic activity.6 These findings suggest that peripheral infiltration and expansion of high-frequency intratumoral T-cell clonotypes are associated with pathological response, but whether they causally drive pCR requires further investigation.6 This suggests that effective antitumor immunity depends on the trafficking of T cells between the peripheral blood and the tumor microenvironment. Consistently, a multimodal sequencing study reported a significantly higher clonal expansion of T cells in the tumors of patients who responded to neoadjuvant nivolumab treatment.25

Single-cell and spatial transcriptomics have provided unprecedented resolution of immune microenvironment remodeling. A pilot study using single-cell-resolution spatial transcriptomics dissected the detailed remodeling of the tumor immune microenvironment following neoadjuvant stereotactic body radiation therapy plus atezolizumab and bevacizumab.26 Post-treatment specimens showed significantly higher anti-cancer immune infiltration, including organized peritumoral aggregates, compared with treatment-naïve HCC specimens.26 Another study employed multimodal sequencing approaches, including spatially-resolved transcriptomics, single-cell and bulk-cell transcriptomics, and TCR clonotype analyses, to examine the spatiotemporal dynamics of neoadjuvant nivolumab treatment in HCC patients.25 The study observed significantly higher clonal expansion of T cells in the tumors of responders, while lipid accumulation was detected in non-responders.25 Furthermore, a functional antagonism between tumor-associated macrophages (TAMs) and CD8 cells was identified, along with a UBASH3B/NR1I2/CEACAM1/HAVCR2 signaling axis highlighting the intense communication among TAMs, tumor cells, and T-cells that leads to pro-tumorigenic outcomes and poorer nivolumab response.25

Multimodal sequencing has further identified synergistic mechanisms driving resistance to neoadjuvant nivolumab treatment, revealing that tumor-associated macrophages, exhausted CD8 T cells, and lipid-enriched tumor cells co-localize at the tumor-immune interface, forming a physical and functional barrier that precludes effective immune cell infiltration into the tumor niche.25

The transition from “cold” to “hot” tumors is a key concept in understanding pCR. Spatial profiling of tumor samples from a neoadjuvant trial combining cabozantinib and nivolumab revealed that tumors from responders were enriched in antigen-presenting cancer-associated fibroblasts expressing HLA-B and CD74, with upregulation of immunomodulatory pathways including TNFα signaling, interferon gamma response, and inflammatory response.3 Responding tumors show enrichment for immune cells and cancer-associated fibroblasts with pro-inflammatory signaling, whereas non-responding tumors exhibit features of immune exclusion and immunosuppression.25 In a pilot study of neoadjuvant stereotactic body radiation therapy plus immunotherapy, one patient achieved pathological complete response, and all resected patients remained relapse-free at data cutoff (median follow-up 16.3 months).26

Immunological Differences Between pCR and Non-pCR Patients

The immunological differences between patients who achieve pCR and those who do not are multifaceted, involving both the tumor and the host immune system.

T-cell repertoire characteristics differ substantially between responders and non-responders. Magen et al demonstrated that response to neoadjuvant PD-1 blockade correlated with the clonal expansion of intratumoral CXCL13+CH25H+IL-21+PD-1+CD4+ T helper cells and Granzyme K+PD-1+ effector-like CD8+ T cells, whereas terminally exhausted CD39hiTOXhiPD-1hiCD8+ T cells dominated in nonresponders.27 Similarly, Zeng et al reported significantly higher clonal expansion of T cells in the tumors of responders to neoadjuvant nivolumab treatment.25 Zhang et al further demonstrated that the peripheral infiltration of high-frequency intratumoral T-cell clonotypes is a critical hallmark of response.6 Collectively, these findings highlight that effective antitumor immunity requires not only the presence of T cells but also their expansion and functional activation.

Myeloid cell composition also differs between responders and non-responders, but these differences are largely correlative. Single-cell and spatial transcriptomic analyses have identified lipid-associated macrophages as mediators of immunosuppression in non-responding tumors. Weng et al demonstrated that macrophage aggregation characterizes a subset of cytotoxic T lymphocyte (CTL)-rich HCCs with an immunosuppressive microenvironment and poor prognosis, and that spatial aggregation enhances pro-tumoral macrophage reprogramming marked by lipid metabolism orientation, M2-like polarization, and increased adjacent CTL exhaustion.28 Furthermore, FABP5+ lipid-loaded macrophages have been shown to process tumor-derived unsaturated fatty acid signals to suppress T-cell antitumor immunity.29 These macrophages may impair T-cell function and contribute to immune evasion.

Stromal features further distinguish responders from non-responders. Spatial transcriptomics profiling demonstrated that the tumor microenvironment of responding tumors was enriched for immune cells and cancer-associated fibroblasts (CAFs) with pro-inflammatory signaling relative to non-responders. Xu et al identified a distinct FMO2+ CAF subset associated with favorable response to anti-PD-1 therapy and better clinical outcomes, which promotes tertiary lymphoid structure formation and increases CD8+ T cell and M1-like macrophage infiltration through the CCL19-CCR7 axis.30 Spatial transcriptomics analysis of neoadjuvant cabozantinib and nivolumab in advanced HCC has further identified independent mechanisms of resistance and recurrence, highlighting the spatial heterogeneity of the tumor microenvironment.31 Cheng et al revealed two fibroblast subsets (CAF-FAP and CAF-C7) with opposing effects on patient prognosis, and identified an intratumor inflammatory hub featured by CAF-FAP plus CD8_PDCD1 proximity and a marginal wound-healing hub with CAF-C7 plus Macrophage_SPP1 co-localization.32 These findings suggest that the stromal compartment may modulate antitumor immune responses, although direct evidence for a causal role in pCR is still limited, and that CAFs change their transcriptional program and intercellular crosstalk according to the spatial context.33

The Role of the Liver Immune Environment

The liver’s unique immune environment has profound implications for neoadjuvant immunotherapy in HCC. The liver is an immunologically privileged organ whose tolerogenic milieu is maintained by multiple hepatic cell types and immunosuppressive cytokines.34 This inherent tolerance poses a major barrier to achieving pCR, as effective neoadjuvant immunotherapy must overcome local immunosuppression to enable tumor-reactive T-cell expansion and tumor eradication.35,36

At the molecular level, hepatic immune tolerance is enforced by a network of immunosuppressive cytokines and checkpoint molecules. LSECs control immune responses by overexpressing inhibitory molecules such as PD-L1 while suppressing co-stimulatory molecules like CD80.37,38 In HCC, Kupffer cells can be co-opted to suppress CD8+ T-cell function through B7-H1/PD-1 interactions, and PD-1+CD8+ T cells in the tumor stroma exhibit reduced proliferation and effector function.39 Importantly, blocking this interaction can restore effector T-cell function.39 Conversely, promoting M1 polarization of Kupffer cells or activating specific Kupffer cell subsets (KC2) can revert T-cell dysfunction and enhance antitumor immunity.13,40 These findings suggest that reversing liver-specific immunosuppression is a key mechanism by which neoadjuvant immunotherapy may drive pCR. Neoadjuvant immunotherapy may thus overcome the liver’s tolerogenic barriers by activating tumor-specific T cells in the presence of the primary tumor, converting the liver from an immunosuppressive to an immune-activating environment, a conversion likely prerequisite for achieving pCR.8 However, the liver immune environment is not uniform; it is significantly shaped by the underlying aetiology of HCC, which may influence responses to ICIs and, consequently, pCR rates.41 HCC arising from viral hepatitis (HBV or HCV) exhibits distinct immune profiles compared with non-viral HCC, which is primarily driven by metabolic dysfunction-associated steatotic liver disease (MASLD).42 In HBV-related HCC, an immune imbalance between CD8+ T cells and Tregs promotes tumor progression, and tumor-reactive T-cell scores and M1-like macrophage signatures are significantly higher than in HBV-negative cases.43 By contrast, MASLD-related HCC is characterized by a more immunosuppressive and exhausted tumor microenvironment.44 Single-cell transcriptomic analyses have revealed that immunosuppressive and exhausted CD8+ T cells are induced during the transition from MASH to HCC.45 In MASLD-related HCC, immune suppression is predominantly driven by metabolic dysfunction rather than viral inflammation, and the lipid-rich microenvironment can induce dendritic cell ferroptosis, impairing CD8+ T-cell responses.46 These aetiology-specific differences may contribute to the heterogeneity of pCR rates across studies and should be considered when selecting patients for neoadjuvant immunotherapy.

In summary, the liver’s tolerogenic environment poses a barrier to pCR, but this barrier can be overcome by effective immunotherapy. The heterogeneity introduced by viral vs metabolic aetiologies further modulates the immune landscape and may explain variable pCR rates, highlighting the need for aetiology-informed patient selection.

Integrated Mechanistic Framework Linking Baseline Immunity to pCR and Recurrence

To integrate the evidence discussed above, we propose a temporal and mechanistic framework that connects baseline immune status, treatment-induced immune remodeling, tumor regression, pCR, and subsequent recurrence or survival (Figure 1).

Step 1: Baseline immune contexture. The pre-treatment tumor immune microenvironment determines the potential for response. High baseline densities of CD8+ TILs and TLS, together with an immune-inflamed phenotype, are associated with a greater likelihood of achieving pCR. Conversely, immune-excluded or immunosuppressive tumors, characterized by lipid-associated macrophages, exhausted CD8+ T cells, and CAF-mediated exclusion, are less likely to respond.

Step 2: Treatment-induced immune remodeling. Neoadjuvant ICIs, alone or in combination, activate tumor-reactive T cells, induce clonal expansion, and promote trafficking of high-frequency intratumoral T-cell clonotypes between peripheral blood and the tumor. This remodeling converts “cold” tumors into “hot” tumors, characterized by increased immune infiltration, formation of TLS, and pro-inflammatory CAF polarization.

Step 3: Tumor regression and pCR. Effective immune remodeling leads to immune-mediated tumor cell killing, resulting in pathological complete response (100% tumor regression). The depth of immune activation, rather than radiological response alone, determines whether pCR is achieved.

Step 4: Post-pCR immune surveillance and recurrence. After pCR, durable antitumor immunity depends on persistent immune surveillance. Recurrence may occur if immune-evasive clones emerge or if residual immunosuppression permits micrometastatic outgrowth. Conversely, sustained immune memory and clearance of ctDNA are associated with prolonged relapse-free survival.

This integrated framework highlights that pCR is not an isolated endpoint but the culmination of a dynamic immune process. Future studies should test this model prospectively by integrating baseline immune profiling, on-treatment immune monitoring, and long-term outcome data.

Biomarkers for Predicting pCR to Neoadjuvant Immunotherapy

Tissue-Based Biomarkers

PD-L1 expression remains the most widely studied tissue biomarker. However, its predictive value for neoadjuvant immunotherapy in HCC is inconsistent, reflecting the complexity of immune checkpoint biology and the limitations of PD-L1 as a standalone biomarker.3 In a cohort of 20 patients receiving neoadjuvant nivolumab, PD-1 expression was associated with treatment response (p = 0.048), whereas PD-L1 expression showed no significant correlation with pathological response (p = 0.26).21 The clinical utility of PD-L1 as a biomarker in HCC is further limited by its heterogeneous expression and dynamic regulation by microenvironmental cues and therapeutic pressure.3 These dynamic changes in PD-L1 expression may reflect the remodeling of the tumor immune microenvironment that ultimately determines pCR.

Tumor-infiltrating lymphocyte (TIL) density is a more robust predictive biomarker compared with PD-L1 expression. As discussed in Baseline Immune Landscape: Predictors of pCR, higher baseline TIL density, particularly CD8+ TILs, correlates with improved pathological responses.22 TILs have also been identified as a prognostic factor.47,48 and a potential predictive biomarker of treatment response in HCC.20 Thus, baseline TIL density may serve not only as a predictive biomarker but also as a mechanistic driver of pCR.

Tertiary lymphoid structures (TLS) have emerged as one of the most promising tissue predictive biomarkers. As noted in Baseline Immune Landscape: Predictors of pCR, TLS are promising predictive biomarkers; high intratumoral TLS density is associated with pathologic response and improved RFS.49 By serving as local hubs for antigen presentation and T-cell activation, TLS may actively promote the immune response required for pCR.

Multiplex immunohistochemistry and spatial profiling enable comprehensive characterization of the immune microenvironment. These approaches can quantify multiple immune cell subsets simultaneously and assess their spatial organization, providing richer predictive information than single-marker approaches. In the PRIME-HCC trial, multiplex immunohistochemistry analysis revealed that pretreatment biopsies of patients with MPR were significantly enriched in peritumoral CD4+ and CD8+ cells.12 Spatial transcriptomics profiling has further identified antigen-presenting cancer-associated fibroblasts (apCAFs) as features associated with response to neoadjuvant immunotherapy in HCC, with spatial analysis revealing that apCAFs in responders are enriched in antigen presentation-related genes and localized in closer proximity to T cells.50 Recent studies have further characterized the spatial and functional polarization of CAF subtypes in HCC, demonstrating that apCAFs are enriched in the peritumoral zone and that the CXCL12-CXCR4 axis may contribute to immune exclusion, highlighting the stromal compartment as an active modulator of antitumor immune responses.50 Spatial immune profiling thus provides mechanistic insights into how immune cell organization drives pCR.

Liquid Biopsy Biomarkers

Circulating tumor DNA (ctDNA) has emerged as a powerful liquid biopsy biomarker for monitoring response to neoadjuvant immunotherapy.51 Baseline levels of cfDNA and ctDNA have been shown to effectively discriminate HCC patients based on their radiological response to immune checkpoint inhibitors.51 Dynamic changes in ctDNA correlate closely with pathological responses and RFS.52 A systematic review and meta-analysis of 13 trials involving 380 patients with detectable ctDNA at baseline demonstrated that ctDNA clearance before or at the time of surgery predicted pathologic complete response with a pooled sensitivity of 0.98 (95% CI 0.86–1.00), though specificity was limited at 0.53 (95% CI 0.37–0.69).52 Postoperative ctDNA clearance is associated with improved 2-year RFS.52 Among patients who achieved pCR, 3-year estimated event-free survival rates ranged from 78.6% to 100%.52

Monitoring ctDNA dynamics during immune-targeted therapy enables real-time assessment of disease status and provides a basis for optimizing treatment strategies.53 A study of 28 HCC patients receiving immune-targeted therapy demonstrated that a decrease in mean variant allele frequency after two cycles of treatment was associated with longer progression-free survival, with a sensitivity of 1.0 and specificity of 0.8125 in predicting partial or complete response, superior to serum alpha-fetoprotein (AFP) levels (sensitivity 0.9167, specificity 0.4667).53

ctDNA has also emerged as a biomarker for minimal residual disease and recurrence surveillance,54 with postoperative minimal residual disease status emerging as one of the most powerful prognostic indicators.52 pCR and ctDNA clearance are increasingly used as intermediate endpoints in neoadjuvant and perioperative immunotherapy trials.52 ctDNA clearance likely reflects effective immune-mediated tumor elimination, linking systemic immune dynamics to pCR.

Serum alpha-fetoprotein (AFP) dynamics may also provide predictive information, though its utility is limited to AFP-producing tumors. Changes in AFP during neoadjuvant treatment may correlate with pathological response. A review of neoadjuvant therapy biomarkers highlighted dynamic serum biomarkers such as AFP kinetics as key advances in predicting treatment efficacy.55 In a multicenter study combining deep learning-assisted radiomic analysis with AFP response, the combined radiomics-AFP model achieved area under the curve values of 0.920 in the test set and 0.857 in the validation set for predicting pathological complete response, demonstrating that integrating AFP dynamics with other biomarkers significantly improves predictive performance.56 Early AFP response has also been shown to be an effective biomarker for identifying high-risk patients prone to early tumor progression and is significantly associated with long-term survival in liver cancer patients treated with immune checkpoint inhibitor combinations.57 AFP decline may reflect immune-mediated tumor lysis, supporting its role as a surrogate of pCR.

Peripheral blood immune cell profiling offers additional predictive information. The peripheral infiltration and expansion of high-frequency intratumoral T-cell clonotypes can be detected in peripheral blood, suggesting that peripheral blood monitoring may provide a non-invasive window into intratumoral immune dynamics. As discussed in Dynamic Changes in the Tumor Immune Microenvironment During Neoadjuvant Immunotherapy, peripheral expansion of high-frequency intratumoral T-cell clonotypes has been associated with pathological response.6 Emerging evidence indicates that the efficacy of immunotherapy hinges not only on reinvigorating pre-existing intratumoral T cells but also on recruiting novel T cells from extratumoral sites, including peripheral blood.6 These findings establish peripheral T-cell clonal expansion as a mechanistic driver of pCR.

Imaging-Based Biomarkers

Radiomics and deep learning approaches are emerging as non-invasive tools for predicting pCR.58 Deep learning-assisted tumor radiomic dynamics on MRI can predict pCR in HCC undergoing immune-based therapy followed by resection.56 A multicenter study developed and validated a model integrating clinicopathological and radiomic features to predict pCR in initially unresectable HCC patients receiving conversion therapy.56 The dynamic radiomic model, which extracted temporal features from baseline, post-treatment, and delta MRIs, significantly outperformed baseline and preoperative models, achieving AUCs of 0.819 in the test set and 0.781 in the validation set for patient-level pCR prediction.56 When combined with AFP response, the radiomics-AFP model achieved even higher AUCs of 0.920 (test) and 0.857 (validation) for lesion-level pCR prediction.56 These approaches analyze quantitative imaging features that reflect tumor heterogeneity, vascularity, and immune infiltration.58 A radiogenomics biomarker integrating genomics with radiology has also shown favorable performance for predicting pathological response and tumor recurrence in a prospective neoadjuvant cohort, accompanied by biological validation through single-cell RNA-seq analysis of pre-treatment biopsies.59

Dynamic contrast-enhanced imaging may capture early changes in tumor perfusion and vascular permeability that correlate with immune activation and subsequent pathological response.56 A multicenter study of 154 patients with initially unresectable HCC demonstrated that deep learning-assisted dynamic radiomic changes on MRI effectively predicted pathological complete response following conversion therapy, with the dynamic radiomic model achieving AUCs of 0.819 in the test set and 0.781 in the validation set for patient-level pCR prediction.56 When combined with AFP response, the integrated model achieved even higher AUCs of 0.920 and 0.857 for lesion-level pCR prediction, significantly outperforming baseline and preoperative models.56 These findings highlight the potential of dynamic imaging biomarkers for non-invasive pCR prediction.56 Changes in tumor size, immune cell infiltration, and immune cell activation have been identified as candidate predictive markers of pathologic response to neoadjuvant immunotherapy in patients with resectable HCC.60 However, the imperfect correlation between radiological response and pathological response highlights the limitations of imaging alone.4 In a cross-trial, patient-level analysis of 111 patients from five Phase I/II trials, while 74% of patients with a radiological response achieved a MPR, 30% of patients with a MPR were not predicted by radiological assessment.4 Furthermore, patients without a MPR had a lower median RFS than patients without a radiological response (28.3 months vs 32.8 months), underscoring the superior prognostic accuracy of pathological assessment over imaging-based response evaluation.4 These findings underscore the importance of integrating imaging with pathological and molecular data to improve the accuracy of pCR prediction and patient selection.55,58 Imaging features reflecting immune infiltration and vascular changes may serve as non-invasive surrogates of the immune dynamics that drive pCR. Table 2 summarizes the current evidence for predictive biomarkers of pCR across tissue, liquid biopsy, and imaging modalities. The integration of tissue, liquid biopsy, and imaging biomarkers for pCR prediction is summarized in Figure 2. Collectively, these biomarkers capture different aspects of the immune dynamics that drive pCR, from baseline immune contexture to treatment-induced remodeling and residual disease clearance. To systematically compare these biomarkers, we have added Table 3, which summarizes their sensitivity, specificity, reproducibility, validation status, and clinical applicability.

Table 2.

Summary of Predictive Biomarkers for Pathological Complete Response to Neoadjuvant Immunotherapy in Hepatocellular Carcinoma

Biomarker Category Biomarker References Predictive Value Key Findings Limitations
Tissue-Based PD-L1 expression [3,34] Inconsistent PD-1 expression associated with response (p=0.048), but PD-L1 not (p=0.26) Heterogeneous expression, dynamic regulation by microenvironmental cues
Tumor-infiltrating lymphocytes (TILs) [14,18,35,36] Promising Higher baseline CD8+ TIL density correlates with improved pathological response; foam cell response and TILs linked to favorable immunotherapy responses Requires standardized quantification methods; intratumoral heterogeneity
Tertiary lymphoid structures (TLS) [15,18,38] Strong High TLS density associated with pathologic response and 100% RFS at 30 months; B-cell depletion abrogates antitumor efficacy in murine models Requires specialized immunohistochemistry; may require multiple biopsy samples
Multiplex immunohistochemistry [11,39] Emerging Enables simultaneous quantification of multiple immune cell subsets; spatial profiling reveals apCAF-T cell proximity in responders High cost; technical expertise required; limited standardization
Liquid Biopsy ctDNA dynamics [13,41] Strong ctDNA clearance predicts pCR with pooled sensitivity 0.98 (95% CI 0.86–1.00); VAFmean decrease predicts response with sensitivity 1.0, specificity 0.8125 Costly; limited accessibility; ctDNA may not be detectable at baseline in all patients
AFP kinetics [43–45] Moderate MPR patients showed median AFP reduction −93.23% vs −5.00% in non-responders; radiomics-AFP model AUC 0.920 for pCR Utility limited to AFP-producing tumors (approximately 60–70% of HCCs)
Peripheral blood T-cell clonotypes [14] Emerging Peripheral infiltration of high-frequency intratumoral T-cell clonotypes is a critical driver of pathological response Requires TCR sequencing; correlation with intratumoral dynamics not fully established
Imaging-Based Radiomics/deep learning [44,46,47] Promising Dynamic radiomic model predicted patient-level pCR with AUC 0.819 (test) and 0.781 (validation); radiomics-AFP model AUC 0.920 Retrospective studies; requires external validation; limited generalizability across different MRI platforms
DCE-MRI [44] Emerging Dynamic contrast-enhanced imaging captures early changes in tumor perfusion correlating with immune activation Preliminary evidence; small sample size studies; requires specialized imaging protocols

Notes: Evidence level reflects the current body of literature and clinical experience: “Strong” indicates consistent results from multiple studies or meta-analyses; “Promising” indicates encouraging results from early-phase or single-center studies; “Moderate” indicates mixed results or limited to specific patient subsets; “Emerging” indicates preliminary evidence requiring further validation. Predictive value is reported as described in the original publications, and assay availability may vary across different clinical settings.

Abbreviations: pCR, pathological complete response; PD-L1, programmed death-ligand 1; PD-1, programmed death-1; TILs, tumor-infiltrating lymphocytes; TLS, tertiary lymphoid structures; RFS, recurrence-free survival; apCAF, antigen-presenting cancer-associated fibroblast; ctDNA, circulating tumor DNA; VAFmean, mean variant allele frequency; CI, confidence interval; AFP, alpha-fetoprotein; MPR, major pathological response; AUC, area under the curve; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging.

Figure 2.

A flow diagram integrating tissue, liquid biopsy and imaging biomarkers into pCR prediction.

Integrated multimodal biomarker model for predicting pCR to neoadjuvant immunotherapy in HCC. Tissue biomarkers (PD-L1, TILs, TLS, multiplex IHC), liquid biopsy biomarkers (ctDNA, AFP, peripheral T-cell clonotypes), and imaging biomarkers (radiomics, DCE-MRI) are integrated through multi-omics approaches to generate pCR probability prediction, guiding patient selection and treatment decisions.

Abbreviations: HCC, hepatocellular carcinoma; pCR, pathological complete response; PD-L1, programmed death-ligand 1; TILs, tumor-infiltrating lymphocytes; TLS, tertiary lymphoid structures; IHC, immunohistochemistry; ctDNA, circulating tumor DNA; AFP, alpha-fetoprotein; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging.

Table 3.

Current Challenges and Future Directions for pCR as an Endpoint in Neoadjuvant Immunotherapy for HCC

Challenge Description Future Directions
Heterogeneity in pCR/MPR definitions While pCR is uniformly defined as 100% tumor regression, MPR thresholds vary between ≥70% and ≥90% tumor regression across studies.4 The D’Alessio et al analysis identified 90% tumor regression as an optimal threshold for predicting improved outcomes, but this has not been universally adopted.4,56 Standardize MPR threshold (≥90% tumor regression) in future trials; establish consensus guidelines through international collaboration4,7,56
Lack of standardized pathological assessment Evaluation of tumor viability requires specialized expertise and is subject to inter-observer variability.18 No consensus exists regarding the pathologic evaluation of HCC specimens after neoadjuvant immunotherapy.18 The absence of central pathology review introduces potential variability57 Develop standardized protocols for specimen processing, sectioning, and assessment;18 leverage digital pathology and artificial intelligence to improve reproducibility across centers;57 implement mandatory central pathology review in pivotal trials7
Limited long-term follow-up data Most studies have reported follow-up durations of 2–3 years.4 pCR remains associated with measurable residual risk of relapse in most tumors, and no event-free survival benefit from postoperative adjuvant immunotherapy was observed among patients achieving pCR in cross-trial comparisons49 Conduct long-term prospective studies with ≥5-year follow-up; establish large-scale patient registries to track real-world outcomes; validate pCR as a surrogate endpoint for overall survival4,49
Small sample sizes in most trials Most neoadjuvant trials are Phase I/II studies with fewer than 50 patients.5,56 Lack of evidence of overall survival benefit in Phase III randomized controlled trials remains a significant gap5,56 Design and conduct adequately powered Phase III trials;5,56 establish multi-center, international consortia to pool data across trials;4 utilize master protocol designs to accelerate evidence generation

Notes: The challenges and future directions summarized in this table are derived from the literature cited in the main text. The proposed future directions represent a synthesis of recommendations from the reviewed studies and expert commentaries. The evidence base for each challenge is supported by the corresponding references, while the future directions reflect emerging consensus in the field.

Abbreviations: pCR, pathological complete response; MPR, major pathological response; HCC, hepatocellular carcinoma; ICI, immune checkpoint inhibitor; RFS, recurrence-free survival.

Prognostic Significance of pCR Following Neoadjuvant Immunotherapy

pCR and Long-Term Survival Outcomes

pCR is consistently associated with improved long-term survival outcomes across multiple studies and tumor types.61 In a pan-tumor reconstructed individual patient analysis of eight solid tumor types including HCC, patients who achieved pCR demonstrated 3-year estimated event-free survival rates ranging from 78.6% to 100% 55. However, these data are derived from a mixed tumor population and may not be directly generalizable to HCC, given differences in tumor biology, treatment regimens, and patient characteristics.61 However, this pan-tumor analysis includes diverse tumor types and treatment regimens, and the applicability to HCC specifically requires further validation. As demonstrated by D’Alessio et al,4 pCR and MPR were associated with significantly improved RFS (HR 0.27 for MPR). These findings were independently highlighted in an accompanying editorial, which noted that pathological response strongly correlates with improved RFS (HR 0.26 for MPR) and has potentially important implications for clinical practice and trial design.7 As reported in Incidence of pCR Across Different Neoadjuvant Immunotherapy Regimens, ICIs significantly increase the pCR rate,9 supporting the clinical relevance of pCR as an efficacy endpoint. The survival benefit associated with pCR is likely driven by the successful remodeling of the tumor immune microenvironment, including expansion of tumor-reactive T-cell clones, formation of tertiary lymphoid structures, and clearance of ctDNA, which collectively establish durable antitumor immunity. HCC-specific prospective studies are therefore needed to confirm the long-term prognostic significance of pCR in this disease.

Collectively, these findings establish pCR as a strong prognostic factor for RFS and support its potential role as an intermediate endpoint. However, its qualification as a surrogate for overall survival remains to be confirmed in prospective trials.4,7,61 The magnitude of the RFS benefit associated with pCR is substantial, suggesting that achieving pCR may be a clinically meaningful goal. Nonetheless, the lack of prospective OS data warrants caution in overinterpreting pCR as a definitive surrogate endpoint.4,7 These immune mechanisms may explain why pCR, rather than radiological response alone, is more strongly associated with long-term outcomes. However, several methodological considerations warrant attention, including the absence of a control group, heterogeneity in treatment regimens, and the lack of standardized pathological assessment across centers.7

pCR and Postoperative Recurrence

Patients achieving pCR have markedly lower rates of postoperative recurrence. In a 25-year retrospective study of 120 HCC patients who underwent liver transplantation, no recurrence or mortality occurred in patients who achieved pathological complete response, establishing pCR as a strong prognostic indicator.62 However, recurrence can still occur even after pCR. In a single-center retrospective study of 20 patients receiving neoadjuvant immunotherapy, one of the ten patients who achieved pCR experienced recurrence during the follow-up period,63 indicating that pCR does not guarantee cure. The mechanisms of recurrence after pCR are not well understood but may involve the emergence of immune-evasive tumor clones or the presence of micrometastatic disease that was not eradicated by neoadjuvant therapy. These mechanisms reflect a failure of immune surveillance and the emergence of immune-evasive clones, underscoring the immune basis of recurrence after pCR and the need for continued immune monitoring.

For patients who do not achieve pCR, the risk of recurrence is significantly higher. For patients who do not achieve pCR, the risk of recurrence is significantly higher, as shown in the cross-trial analysis by D’Alessio et al4 This highlights the need for adjuvant strategies in patients with incomplete pathological responses. The prognostic value of pathological response has been further validated in the conversion therapy setting, where pCR was associated with improved overall survival (HR 0.29) and RFS (HR 0.40).64 The higher recurrence risk in non-pCR patients may reflect incomplete immune remodeling and persistent immunosuppression within the tumor microenvironment.

pCR in Special Populations

Liver transplantation represents a unique context for pCR assessment. Early-phase trials and multicenter cohorts demonstrate that pre-LT ICIs achieve downstaging in approximately 75% to 82% of patients, with radiologic responses up to 94% and pathologic responses ranging from 35% to 88%.65 One- and three-year survival after ICI bridging reached approximately 95% and 70% to 80%, while post-LT survival remained above 85% at 3 years.65 pCR in this setting may allow patients with initially unresectable or transplant-ineligible HCC to undergo curative transplantation. However, the risk of graft rejection is substantial, with acute rejection episodes occurring in 16% to 28% of patients across large series.65 The duration of the washout interval between the last ICI dose and transplantation is a critical determinant of rejection risk, with intervals shorter than 30 to 90 days correlating with markedly elevated rejection rates.65 In a cohort of 10 patients with intermediate-advanced HCC receiving pre-LT ICI-based downstaging therapy, 4 patients demonstrated complete pathological response, and the 12- and 24-month overall survival rates post-LT were 75%.66 In this setting, pCR reflects a balance between effective immune activation and the risk of immune-mediated graft rejection, highlighting the need to modulate immune dynamics carefully. These findings highlight the importance of careful patient selection, strict washout protocols, and individualized management of immunosuppression.65

Unlike neoadjuvant therapy for initially resectable HCC, conversion therapy is applied in patients with initially unresectable disease and aims to achieve secondary resectability. Its distinct patient population and treatment goals should be considered when interpreting pCR rates. Unresectable HCC conversion therapy represents another important application. Neoadjuvant immunotherapy can downstage locally advanced tumors, converting initially unresectable disease into resectable disease. In the GUIDANCE004 study of 1262 patients with initially unresectable HCC who received immune-based conversion therapy, pCR was associated with better overall survival (HR 0.29) and RFS (HR 0.40).64 These survival benefits were confirmed in landmark analyses and propensity score-matched subgroups.64 Among those undergoing hepatectomy after conversion therapy, radiographic complete response (rCR) was strongly associated with higher pCR rates (67.1% vs 13.8% in non-rCR patients; Φ = 0.55), and rCR was identified as the only preoperative predictor of pCR (odds ratio 16.65).64 In a single-arm exploratory trial of hepatic arterial infusion chemotherapy combined with camrelizumab and apatinib as conversion therapy, 22.2% (2 of 9) of patients who underwent surgery achieved pathological complete response.67 pCR in this setting is associated with particularly favorable outcomes, as these patients often have more advanced disease at baseline. This survival advantage may be attributable to conversion of an immune-excluded tumor microenvironment into an immune-inflamed one, enabling durable immune-mediated tumor control. A clinical decision framework integrating pCR assessment into postoperative management is proposed in Figure 3.

Figure 3.

Flowchart sequence: Resectable HCC patient, MDT evaluation, neoadjuvant ICI decision, pCR decision, outcomes. Flowchart sequence: Resectable HCC patient, MDT evaluation, Suitable for neoadjuvant ICI, Neoadjuvant ICI therapy, Imaging plus ctDNA assessment, pCR achieved, Surgery, Long-term follow-up or surveillance, Partial response, Surgery plus consider adjuvant, No response, Alternative therapy.

Clinical decision flowchart for resectable HCC following neoadjuvant immunotherapy based on pCR. After MDT evaluation, eligible patients receive 2–3 cycles of ICI therapy. Response is assessed by imaging and ctDNA. Based on pCR status, patients are triaged into three pathways: pCR to surgery; partial response to surgery with consideration of adjuvant therapy; no response to alternative therapy. All patients undergo long-term follow-up.

Abbreviations: HCC, hepatocellular carcinoma; pCR, pathological complete response; MDT, multidisciplinary team; ICI, immune checkpoint inhibitor; ctDNA, circulating tumor DNA.

Challenges, Limitations, and Future Directions

Current Challenges and Limitations

Several challenges currently limit the clinical application of pCR as an endpoint and the optimization of neoadjuvant immunotherapy in HCC.

Heterogeneity in pCR and MPR definitions remains a significant obstacle. While pCR is uniformly defined as 100% tumor regression, MPR thresholds vary between ≥70% and ≥90% tumor regression in different studies. As discussed in pCR as a Surrogate Endpoint: Evidence and Controversies, the optimal MPR threshold remains debated.4,68 This definitional heterogeneity complicates cross-trial comparisons and meta-analyses.

Lack of standardized pathological assessment is another challenge. The evaluation of tumor viability in resected specimens requires expertise and may be subject to inter-observer variability. A comprehensive review emphasized that for HCC specimens undergoing neoadjuvant or conversion ICI-related therapy, it is urgent to establish a feasible strategy to appraise and quantify immunotherapy-related pathological response.20 Although there is no consensus regarding the pathologic evaluation of relevant HCC specimens, several studies have begun to focus on this field. The absence of central pathological review in many studies introduces potential variability.69 Recent advances in digital pathology and artificial intelligence may help standardize response assessment and improve reproducibility across different centers.69

Limited long-term follow-up data constrain the definitive qualification of pCR as a surrogate endpoint. Most studies have reported follow-up durations of 2–3 years. A pan-tumor reconstructed individual patient analysis of 1867 patients achieving pCR across 8 tumor types demonstrated that 3-year estimated event-free survival rates ranged from 78.6% to 100%, but pCR remained associated with measurable residual risk of relapse in most tumors 55. These findings may not be directly applicable to HCC, and HCC-specific validation is still lacking.61 In cross-trial comparisons, no event-free survival benefit from postoperative adjuvant immunotherapy was observed among patients achieving pCR.61 Most importantly, no prospective randomized trial has yet demonstrated that pCR is a valid surrogate for overall survival in HCC.

Small sample sizes in most neoadjuvant trials limit the statistical power for subgroup analyses and biomarker discovery. Most studies are Phase I/II trials with fewer than 50 patients. A meta-analysis of neoadjuvant immunotherapy in resectable HCC reported a pooled pCR rate of 22%, but the field still faces challenges such as lack of evidence of overall survival benefit in Phase III randomized controlled trials, treatment-related adverse reactions that may delay surgery, and the need for optimal population screening and timing of surgery.5,68 Larger Phase III trials with definitive support for this approach are needed to justify its clinical application. Table 4 summarizes the current challenges and proposed future directions for the clinical application of pCR as an endpoint.

Table 4.

Systematic Comparison of Predictive Biomarkers for pCR to Neoadjuvant Immunotherapy in HCC

Biomarker Sensitivity Specificity Reproducibility Validation Cohorts Clinical Applicability
TILs Not systematically reported; correlates with response Not systematically reported Requires standardized quantification; inter-observer variability Mostly retrospective, single-center Requires pathological expertise; not yet ready for routine use
TLS High in some studies (associated with 100% RFS at 30 months) Not systematically reported Requires specialized IHC; may need multiple biopsies Limited prospective validation Promising but not standardized
ctDNA 0.98 (95% CI 0.86–1.00) for pCR 0.53 (95% CI 0.37–0.69) Requires standardized assays; limited accessibility Pooled analysis of 13 trials (380 patients) Minimally invasive; costly; not universally available
AFP 0.9167 in one study 0.4667 in one study Widely available; reproducible Multicenter retrospective Limited to AFP-producing tumors (~60–70% of HCC)
Radiomics / deep learning AUC 0.819 (test), 0.781 (validation) Not reported as sensitivity/specificity Requires standardized imaging protocols; platform-dependent Multicenter retrospective; external validation limited Non-invasive; requires specialized software and expertise
Immune-cell profiling (peripheral T-cell clonotypes) Not systematically reported Not systematically reported Requires TCR sequencing; correlation with intratumoral dynamics not fully established Early-phase studies Research setting only; not yet clinically applicable

Abbreviations: pCR, pathological complete response; TILs, tumor-infiltrating lymphocytes; TLS, tertiary lymphoid structures; ctDNA, circulating tumor DNA; AFP, alpha-fetoprotein; AUC, area under the curve; IHC, immunohistochemistry; TCR, T-cell receptor; RFS, recurrence-free survival.

Future Directions

Standardization of pCR assessment is an urgent priority.20,69 Consensus guidelines for pathological assessment of HCC in the immunotherapy era are needed. These guidelines should address specimen handling, definition of tumor viability, and reporting standards.20,69 The identification of 90% tumor regression as an optimal threshold in the D’Alessio et al analysis provides a foundation for standardization, but this threshold requires validation in prospective studies and endorsement by international consensus panels.4,7 Recent advances in digital pathology and artificial intelligence may facilitate reproducible and quantitative assessment of tumor viability across centers.69

Multi-omics integration holds promise for improving pCR prediction.25,50 Integrating genomic, transcriptomic, proteomic, and radiomic data may enable more accurate patient stratification.25,50 Single-cell and spatial transcriptomics are already providing unprecedented insights into the immune dynamics of response and resistance.26,27 The integration of multimodal data from large patient cohorts through machine learning approaches may further enhance predictive accuracy and identify novel biomarkers.59 The radiomics-AFP model achieving AUCs of 0.920 for pCR prediction exemplifies the potential of combining multi-omics data with clinical parameters.56

Biomarker-guided patient selection is essential for optimizing neoadjuvant immunotherapy.4 Identifying patients most likely to achieve pCR can maximize the benefit-to-risk ratio of neoadjuvant therapy.4 Baseline TIL density,20 TLS presence,23,24 and ctDNA dynamics52,53 are among the most promising biomarkers. The high sensitivity (0.98) of ctDNA clearance for predicting pCR suggests that ctDNA dynamics could serve as an early decision-making tool for treatment modification.52 Future research should focus on developing validated, clinically applicable biomarker panels that integrate multiple complementary biomarkers to improve predictive performance.

Optimization of combination strategies is needed to improve pCR rates while minimizing toxicity.12 Future studies should specifically investigate how combination strategies can overcome poor immune infiltration and immunosuppression to enhance pCR rates, as suggested by recent evidence.18 Dual ICI combinations achieve higher pCR rates up to 41% but carry greater toxicity, with severe treatment-related adverse events occurring in 41% of patients.12 Identifying the optimal combinations, sequences, and durations for different patient populations is a key research priority.68 The integration of novel agents, including anti-LAG-3, anti-TIGIT, and other immunomodulatory antibodies, into neoadjuvant regimens may further enhance antitumor efficacy, though careful evaluation of their safety profiles is warranted.70,71

Large-scale prospective validation is critical.4,5,61 Phase III trials with adequate sample sizes and long-term follow-up are needed to definitively validate pCR as a surrogate endpoint and to establish evidence-based guidelines for neoadjuvant immunotherapy in HCC.4,5,61 International collaborative efforts to establish patient registries and pool data across trials will accelerate evidence generation and enable more robust subgroup analyses.4 International collaborative efforts to establish patient registries and pool data across trials will accelerate evidence generation and enable more robust subgroup analyses. A matched analysis of the 2017–2022 National Cancer Database demonstrated that neoadjuvant immunotherapy is associated with improved survival after hepatectomy in HCC patients, highlighting the value of large-scale registry-based studies in generating real-world evidence.72 Future studies should prospectively validate the integrated mechanistic framework proposed in Integrated Mechanistic Framework Linking Baseline Immunity to pCR and Recurrence, using longitudinal immune profiling and long-term outcome data.

Conclusion

pCR following neoadjuvant immunotherapy in hepatocellular carcinoma (HCC) represents a promising but not yet fully validated surrogate endpoint. While pCR is consistently associated with improved relapse-free survival (RFS), no prospective randomized trial has yet demonstrated that achieving pCR directly translates into improved overall survival (OS). This distinction should guide the interpretation of pCR in both clinical practice and trial design.

The immune dynamics underlying pCR reflect a continuous and dynamic process rather than an isolated event. Baseline immune contexture, particularly CD8+ T-cell infiltration and tertiary lymphoid structure (TLS) formation sets the stage for response. Effective neoadjuvant immunotherapy then induces T-cell clonal expansion, peripheral trafficking, and remodeling of the tumor microenvironment from an immune-excluded to an immune-inflamed phenotype. These changes collectively drive tumor regression and, ultimately, pCR. This integrated framework highlights the importance of evaluating immune remodeling longitudinally rather than relying on single static biomarkers.

Clinically, pCR offers a valuable early indicator of treatment efficacy and may inform patient selection and postoperative management. However, several unresolved issues limit its definitive qualification as a surrogate endpoint. These include heterogeneity in pathological assessment and MPR thresholds (rather than in the definition of pCR itself), the lack of standardized central pathology review, limited long-term follow-up data, and small sample sizes in most trials. These gaps must be addressed before pCR can be universally adopted as a definitive endpoint.

Future directions should prioritize prospective validation of pCR as a surrogate for OS, standardization of pathological assessment protocols, integration of multi-omics and liquid biopsy biomarkers for improved prediction, optimization of combination strategies to overcome immune suppression, and large-scale collaborative efforts to pool data across trials. Multidisciplinary collaboration among interventional radiologists, hepatobiliary surgeons, medical oncologists, and pathologists will be essential to translate these advances into improved outcomes for patients with HCC.

Abbreviations

AFP, alpha-fetoprotein; apCAFs, antigen-presenting cancer-associated fibroblasts; AUC, area under the curve; CAFs, cancer-associated fibroblasts; CI, confidence interval; ctDNA, circulating tumor DNA; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging; HCC, hepatocellular carcinoma; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IHC, immunohistochemistry; MPR, major pathological response; OR, odds ratio; OS, overall survival; pCR, pathological complete response; PD-1, programmed death-1; PD-L1, programmed death-ligand 1; RFS, recurrence-free survival; TILs, tumor-infiltrating lymphocytes; TKIs, tyrosine kinase inhibitors; TLS, tertiary lymphoid structures; TME, tumor microenvironment.

Disclosure

The authors report no conflicts of interest in this work.

References

  • 1.Malik AK, Geh D, Jeffry Evans TR, et al. Improving surgical treatments for hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2026;23(3):208–19. doi: 10.1038/s41575-025-01143-y [DOI] [PubMed] [Google Scholar]
  • 2.Akula V, Chen L, Acikgoz Y, et al. Neoadjuvant immune checkpoint inhibitors for hepatocellular carcinoma. NPJ Precis Oncol. 2025;9(1):60. doi: 10.1038/s41698-025-00846-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Llovet JM, Pinyol R, Yarchoan M, et al. Adjuvant and neoadjuvant immunotherapies in hepatocellular carcinoma. Nat Rev Clin Oncol. 2024;21(4):294–311. doi: 10.1038/s41571-024-00868-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.D’alessio A, Stefanini B, Blanter J, et al. Pathological response following neoadjuvant immune checkpoint inhibitors in patients with hepatocellular carcinoma: a cross-trial, patient-level analysis. The Lancet Oncol. 2024;25(11):1465–1475. doi: 10.1016/S1470-2045(24)00457-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Cicerone O, Oliviero B, Mantovani S, et al. Neoadjuvant immunotherapy in resectable hepatocellular carcinoma: a meta-analysis of the current evidence. World J Clin Oncol. 2025;16(10):110511. doi: 10.5306/wjco.v16.i10.110511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zhang Y, Tsui Y, Lee JM, et al. Spatiotemporal dynamics of t cells in peripheral blood and tumor underlying differential responses to neoadjuvant pd-1 blockade in hepatocellular carcinoma. J Immun Cancer. 2025;13(12):13. doi: 10.1136/jitc-2025-013516 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Cai L, Chen J, Xu J, et al. Neoadjuvant immunotherapy for hepatocellular carcinoma: progress and perspectives. Hepatobiliary Surg Nutr. 2025;14(1):140–142. doi: 10.21037/hbsn-2024-642 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Tian C, Yu Y, Wang Y, et al. Neoadjuvant immune checkpoint inhibitors in hepatocellular carcinoma: a meta-analysis and systematic review. Front Immunol. 2024;15:1352873. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tuersun A, Mohetaer M, Hou G, et al. Meta-analysis of the efficiency and safety of neoadjuvant therapy with immune checkpoint inhibitors in resectable hepatocellular carcinoma. Front Med. 2024;11:1511511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Zhao M, Chen S, Li C, et al. Neoadjuvant immune checkpoint inhibitors for resectable hepatocellular carcinoma: a systematic review and meta-analysis. Cancers. 2023;16(1):15. doi: 10.3390/cancers16010015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Yuza K, Akabane M, Kawashima J, et al. Hepatocellular carcinoma: new therapies on the horizon. Expert Rev Gastroenterol Hepatol. 2026;20(7):657–672. doi: 10.1080/17474124.2026.2702585 [DOI] [PubMed] [Google Scholar]
  • 12.Pinato DJ, Cortellini A, Sukumaran A, et al. Prime-hcc: phase ib study of neoadjuvant ipilimumab and nivolumab prior to liver resection for hepatocellular carcinoma. BMC Cancer. 2021;21(1):301. doi: 10.1186/s12885-021-08033-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.De Simone G, Andreata F, Bleriot C, et al. Identification of a kupffer cell subset capable of reverting the t cell dysfunction induced by hepatocellular priming. Immunity. 2021;54(9):2089–2100. doi: 10.1016/j.immuni.2021.05.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Wang Z, Fan J, Zhou S, et al. Perioperative camrelizumab plus rivoceranib versus surgery alone in patients with resectable hepatocellular carcinoma at intermediate or high risk of recurrence (cares-009): a randomised phase 2/3 trial. Lancet. 2025;406(10515):2089–2099. doi: 10.1016/S0140-6736(25)01720-9 [DOI] [PubMed] [Google Scholar]
  • 15.Lin Y, Ou D, Su Y, et al. Nivolumab plus ipilimumab for potentially resectable hepatocellular carcinoma: long-term efficacy and biomarker exploration. J Hepatol. 2026;84(2):316–328. doi: 10.1016/j.jhep.2025.08.035 [DOI] [PubMed] [Google Scholar]
  • 16.Dai Q, Deng H, Wang J, et al. Neoadjuvant hepatic arterial infusion chemotherapy with folfox plus tislelizumab for resectable hepatocellular carcinoma beyond the milan criteria: efficacy, safety and biomarker exploration. J Hepatocell Carcinoma. 2026;13:615120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Chen L, Han R, Zhai S, et al. Perioperative tislelizumab for early-stage hepatocellular carcinoma: a phase ii trial with integrated tumour microenvironment profiling and predictive modeling. Drug Resist Updat. 2026;88:101440. doi: 10.1016/j.drup.2026.101440 [DOI] [PubMed] [Google Scholar]
  • 18.Tong J, Tan Y, Ouyang W, et al. Targeting immune checkpoints in hepatocellular carcinoma therapy: toward combination strategies with curative potential. Exp Hematol Oncol. 2025;14(1):65. doi: 10.1186/s40164-025-00636-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Ou-Yang L, Gan Y, Zhang C, et al. Pathological response >/=90% predicts survival in hcc treated with neoadjuvant loco-regional therapies: a retrospective cohort study. JHEP Rep. 2026;8(1):101638. doi: 10.1016/j.jhepr.2025.101638 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wang H, Qian Y, Dong H, et al. Pathologic assessment of hepatocellular carcinoma in the era of immunotherapy: a narrative review. Hepatobiliary Surg Nutr. 2024;13(3):472–493. doi: 10.21037/hbsn-22-527 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Cheung T, Wai-Hung Ho D, Lyu SX, et al. Multimodal integrative genomics and pathology analyses in neoadjuvant nivolumab treatment for intermediate and locally advanced hepatocellular carcinoma. Liver Cancer. 2024;13(1):70–88. doi: 10.1159/000531176 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Marron TU, Fiel MI, Hamon P, et al. Neoadjuvant cemiplimab for resectable hepatocellular carcinoma: a single-arm, open-label, phase 2 trial. Lancet Gastroenterol Hepatol. 2022;7(3):219–229. doi: 10.1016/S2468-1253(21)00385-X [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Wu Q, Jin L, Cao L, et al. Tertiary lymphoid structures (tls): the mechanism hub and therapeutic target of neoadjuvant immunotherapy for hepatocellular carcinoma. Crit Rev Oncol Hematol. 2026;224:105393. [DOI] [PubMed] [Google Scholar]
  • 24.Shu DH, Ho WJ, Kagohara LT, et al. Immune landscape of tertiary lymphoid structures in hepatocellular carcinoma (hcc) treated with neoadjuvant immune checkpoint blockade. bioRxiv. 2023.
  • 25.Zeng F, Zhang Q, Tsui Y, et al. Multimodal sequencing of neoadjuvant nivolumab treatment in hepatocellular carcinoma reveals cellular and molecular immune landscape for drug response. Mol Cancer. 2025;24(1):110. doi: 10.1186/s12943-025-02314-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Cosner Z, Guo Z, Nawrocki C, et al. Neoadjuvant stereotactic body radiotherapy plus immune therapy favorably remodels the hepatocellular carcinoma tumor microenvironment. Clin Cancer Res. 2026;32(14):2885–2897. doi: 10.1158/1078-0432.CCR-25-4779 [DOI] [PubMed] [Google Scholar]
  • 27.Magen A, Hamon P, Fiaschi N, et al. Intratumoral dendritic cell-cd4(+) t helper cell niches enable cd8(+) t cell differentiation following pd-1 blockade in hepatocellular carcinoma. Nat Med. 2023;29(6):1389–1399. doi: 10.1038/s41591-023-02345-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Weng Y, Wang L, Wang Y, et al. Spatial organization of macrophages in ctl-rich hepatocellular carcinoma influences ctl antitumor activity. Cancer Immunol Res. 2025;13(3):310–322. doi: 10.1158/2326-6066.CIR-24-0589 [DOI] [PubMed] [Google Scholar]
  • 29.Yang X, Deng B, Zhao W, et al. Fabp5(+) lipid-loaded macrophages process tumour-derived unsaturated fatty acid signal to suppress t-cell antitumour immunity. J Hepatol. 2025;82(4):676–689. doi: 10.1016/j.jhep.2024.09.029 [DOI] [PubMed] [Google Scholar]
  • 30.Xu W, Weng J, Zhao Y, et al. Fmo2(+) cancer-associated fibroblasts sensitize anti-pd-1 therapy in patients with hepatocellular carcinoma. J Immun Cancer. 2025;(5):13. doi: 10.1136/jitc-2025-011648 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zhang S, Yuan L, Danilova L, et al. Spatial transcriptomics analysis of neoadjuvant cabozantinib and nivolumab in advanced hepatocellular carcinoma identifies independent mechanisms of resistance and recurrence. Genome Med. 2023;15(1):72. doi: 10.1186/s13073-023-01218-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Cheng Y, Chen X, Feng L, et al. Stromal architecture and fibroblast subpopulations with opposing effects on outcomes in hepatocellular carcinoma. Cell Discov. 2025;11(1):1. doi: 10.1038/s41421-024-00747-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Liu Y, Dong G, Yu J, et al. Integration of single-cell and spatial transcriptomics reveals fibroblast subtypes in hepatocellular carcinoma: spatial distribution, differentiation trajectories, and therapeutic potential. J Transl Med. 2025;23(1):198. doi: 10.1186/s12967-025-06192-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Tabrizian P, Marino R, Chow PKH. Liver resection and transplantation in the era of checkpoint inhibitors. JHEP Rep. 2024;6(11):101181. doi: 10.1016/j.jhepr.2024.101181 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Altalbawy FMA, Zwamel AH, Sanghvi G, et al. Interactions in hepatic tumor microenvironment: potential targets and modulations for effective therapy. Pathol Res Pract. 2025;272:156074. [DOI] [PubMed] [Google Scholar]
  • 36.Zhou J, Wang J, Zhang L, et al. Defenestration of liver sinusoidal endothelial cells: the trigger of liver fibrosis. Pharmaceuticals. 2025;18(6):18. doi: 10.3390/ph18060893 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Devan AR, Kumar AR, Nair B, et al. Insights into an immunotherapeutic approach to combat multidrug resistance in hepatocellular carcinoma. Pharmaceuticals. 2021;14(7):656. doi: 10.3390/ph14070656 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Cogliati B, Yashaswini CN, Wang S, et al. Friend or foe? The elusive role of hepatic stellate cells in liver cancer. Nat Rev Gastroenterol Hepatol. 2023;20(10):647–661. doi: 10.1038/s41575-023-00821-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Wu K, Kryczek I, Chen L, et al. Kupffer cell suppression of cd8+ t cells in human hepatocellular carcinoma is mediated by b7-h1/programmed death-1 interactions. Cancer Res. 2009;69(20):8067–8075. doi: 10.1158/0008-5472.CAN-09-0901 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Liu N, Wang X, Steer CJ, et al. Microrna-206 promotes the recruitment of cd8(+) t cells by driving m1 polarisation of kupffer cells. Gut. 2022;71(8):1642–1655. doi: 10.1136/gutjnl-2021-324170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Pfister D, Nunez NG, Pinyol R, et al. Nash limits anti-tumour surveillance in immunotherapy-treated hcc. Nature. 2021;592(7854):450–456. doi: 10.1038/s41586-021-03362-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Cheu JW, Wong CC. The immune microenvironment of steatotic hepatocellular carcinoma: current findings and future prospects. Hepatol Commun. 2024;8(9):8. doi: 10.1097/HC9.0000000000000516 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Zhang S, Xu H, Li M, et al. An etiology-stratified single-cell atlas identifies fabp4 as a prognostic marker for masld-related hcc. J Hepatol. 2026;84(5):920–932. doi: 10.1016/j.jhep.2025.10.026 [DOI] [PubMed] [Google Scholar]
  • 44.Shi Y, Taherifard E, Saeed A, et al. Masld-related hcc: a comprehensive review of the trends, pathophysiology, tumor microenvironment, surveillance, and treatment options. Curr Issues Mol Biol. 2024;46(6):5965–5983. doi: 10.3390/cimb46060356 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Huang Y, Xie Y, Zhang Y, et al. Single-cell transcriptome reveals the reprogramming of immune microenvironment during the transition from mash to hcc. Mol Cancer. 2025;24(1):177. doi: 10.1186/s12943-025-02370-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Li N, Song X, Peng X, et al. Tim-3 facilitates dendritic cell ferroptosis and impairs antitumor immunity in steatohepatitis-related hcc. J Hepatol. 2026;85(2):331–344. doi: 10.1016/j.jhep.2026.04.010 [DOI] [PubMed] [Google Scholar]
  • 47.Pang Y, Zhao X, Guo X, et al. An early relapse prediction model based on pathological features following neoadjuvant immunotherapy for hepatocellular carcinoma. Oncologist. 2025;31(1):31. doi: 10.1093/oncolo/oyaf368 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.He X, Liu Y, Dai T, et al. Pathological study of the tumor microenvironment after neoadjuvant therapy in hepatocellular carcinoma: difference of tace combined with antiangiogenics and immunotherapy. Hepatol Commun. 2025;9(9):9. doi: 10.1097/HC9.0000000000000787 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Shu DH, Ho WJ, Kagohara LT, et al. Immunotherapy response induces divergent tertiary lymphoid structure morphologies in hepatocellular carcinoma. Nat Immunol. 2024;25(11):2110–2123. doi: 10.1038/s41590-024-01992-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Nakamura D, Nishio T, Oguma Y, et al. Spatial and functional polarization of cancer-associated fibroblasts with cxcr4-mediated immune modulation in hepatocellular carcinoma. Hepatol Commun. 2026;10(4):10. doi: 10.1097/HC9.0000000000000930 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Vargas-Accarino E, Higuera M, Bermudez-Ramos M, et al. Harnessing plasma biomarkers to predict immunotherapy outcomes in hepatocellular carcinoma: the role of cfdna, ctdna, and cytokines. Int J Mol Sci. 2025;26(6):26. doi: 10.3390/ijms26062794 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Valenza C, Saldanha EF, Gong Y, et al. Circulating tumor dna clearance as a predictive biomarker of pathologic complete response in patients with solid tumors treated with neoadjuvant immune checkpoint inhibitors: a systematic review and meta-analysis. Ann Oncol. 2025;36(7):726–736. doi: 10.1016/j.annonc.2025.03.019 [DOI] [PubMed] [Google Scholar]
  • 53.Zheng S, Zhang D, Wu J, et al. Circulating tumor dna in peripheral blood may predict the efficacy of immune-targeted therapy in patients with hepatocellular carcinoma. Clin Res Hepatol Gastroenterol. 2025;49(7):102632. doi: 10.1016/j.clinre.2025.102632 [DOI] [PubMed] [Google Scholar]
  • 54.Xu J, Yang L, Wang S, et al. Hepatobiliary surgery in the era of immunotherapy: integrating surgical timing, immune surveillance, and recurrence prevention. Front Immunol. 2026;17:1865909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Palihati S, Mo H, Ye X, et al. Neoadjuvant therapy for hepatocellular carcinoma: emerging combinations and biomarker challenges. Front Immunol. 2026;17:1809834. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Zhou S, Wang L, Wu L, et al. Deep learning-assisted tumor radiomic dynamics on mri predict pathological complete response in hcc undergoing immune-based therapy followed by hepatectomy. Hepatology. 2026;84(2):453–468. doi: 10.1097/HEP.0000000000001724 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Tian J, Pan S, Wang Y, et al. Early alpha-fetoprotein response predicts sustained tumor response following immune checkpoint inhibitors combined with targeted therapy in liver cancer. Biomedicines. 2024;12(12):12. doi: 10.3390/biomedicines12122769 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Miranda J, Horvat N, Fonseca GM, et al. Current status and future perspectives of radiomics in hepatocellular carcinoma. World J Gastroenterol. 2023;29(1):43–60. doi: 10.3748/wjg.v29.i1.43 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Xu Z, Liu Y, Ji Y, et al. Radiogenomics predicts immune microenvironment heterogeneity and response to combination immunotherapy in hepatocellular carcinoma. J Transl Med. 2026;24(1):181. doi: 10.1186/s12967-025-07627-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Lapelusa M, Chamseddine S, Tran Cao HS, et al. Tissue and imaging biomarkers of response to neoadjuvant nivolumab or nivolumab plus ipilimumab in patients with resectable hepatocellular carcinoma. Oncology. 2025;103(6):490–497. doi: 10.1159/000541250 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Garitaonaindia Y, Witteveen S, Dionisio E, et al. Prognostic dynamics of pathological complete response and ctdna clearance after neoadjuvant/perioperative immunotherapy in cancer: a reconstructed individual patient analysis. J Natl Cancer Inst. 2026:djag197. [DOI] [PubMed] [Google Scholar]
  • 62.Agalar C, Egeli T, Ozbilgin M, et al. Pathologic complete response as a prognostic indicator in bridging therapy: a 25-year retrospective study on hepatocellular carcinoma and liver transplantation outcomes. Transpl Proc. 2025;57(9):1768–1772. doi: 10.1016/j.transproceed.2025.08.004 [DOI] [PubMed] [Google Scholar]
  • 63.Podrascanin V, Ammann M, Dong Y, et al. Liver resection following neoadjuvant immunotherapy: a single center retrospective study. Updates Surg. 2026;78(1):219–227. doi: 10.1007/s13304-025-02310-1 [DOI] [PubMed] [Google Scholar]
  • 64.Yang D, Yan Y, Lai Y, et al. Prognostic value of radiological and pathological complete response following immune-based conversion therapy in patients with unresectable hepatocellular carcinoma (guidance004). JHEP Rep. 2025;7(11):101587. doi: 10.1016/j.jhepr.2025.101587 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Khalaf MH, Tran B, Krendl F, et al. Hepatocellular carcinoma and liver transplant: what about neo- and adjuvant immunotherapy. Hepato Pancreat Dis Int. 2026;25(3):270–279. doi: 10.1016/j.hbpd.2025.10.004 [DOI] [PubMed] [Google Scholar]
  • 66.Cheng C, Shi H, Lai Y, et al. Pretransplant immune checkpoint inhibitor treatment for intermediate-advanced hepatocellular carcinoma in living donor liver transplantation. Transpl Direct. 2025;11(10):e 1852. doi: 10.1097/TXD.0000000000001852 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Yalikun K, Li Z, Zhang J, et al. Hepatic artery infusion chemotherapy combined with camrelizumab and apatinib as conversion therapy for patients with unresectable hepatocellular carcinoma: a single-arm exploratory trial. BMC Cancer. 2025;25(1):838. doi: 10.1186/s12885-025-14250-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Yin Z, Chen D, Liang S, et al. Neoadjuvant therapy for hepatocellular carcinoma. J Hepatocell Carcinoma. 2022;9:929–946. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Wen H, Liang R, Liu X, et al. Predicting pathological response of neoadjuvant conversion therapy for hepatocellular carcinoma patients using ct-based radiomics model. J Hepatocell Carcinoma. 2024;11:2145–2157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.El-Kafrawy SA, Elkafrawy MS, Azhar EI, et al. Antibody treatment of hepatocellular carcinoma: a review of current and emerging approaches. Front Immunol. 2025;16:1533874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71.Arvanitakis K, Germanidis G. Immunity revealed: stratifying hepatocellular carcinoma for precision immunotherapy. Mol Ther Oncol. 2025;33(1):200939. doi: 10.1016/j.omton.2025.200939 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Chiang L, Bodzin A, Lin D, et al. Neoadjuvant immunotherapy for hepatocellular carcinoma is associated with improved survival after hepatectomy: a matched analysis of the 2017 to 2022 national cancer database. Cancers. 2026;18(13):18. doi: 10.3390/cancers18132137 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Hepatocellular Carcinoma are provided here courtesy of Dove Press

RESOURCES