Precision targeting of pathological angiogenesis in liver cirrhosis: Molecular mechanisms and therapeutic translation challenges

Abstract

The core lethal complications of liver cirrhosis, particularly portal hypertension and hepatic decompensation, are closely linked to dysregulated pathological angiogenesis. Aberrant neovascularization is primarily driven by the hypoxia-inducible factor-vascular endothelial growth factor (VEGF) axis. Activated hepatic stellate cells release key factors like VEGF and platelet-derived growth factor, promoting endothelial cell proliferation. Inflammatory-oxidative stress signals (e.g., tumor necrosis factor-alpha/interleukin-6 and non-alcoholic fatty liver disease oxidase-derived reactive oxygen species) amplify this process via the nuclear factor kappa-B pathway. Activation of the local renin-angiotensin system also exacerbates vascular leakage and malformation through pathways like angiotensin II. Preclinical studies demonstrate that drugs targeting these pathways (e.g., the anti-VEGF agent bevacizumab, the multi-kinase inhibitor sorafenib) can effectively inhibit pathological angiogenesis. However, their clinical translation faces three major challenges: (1) Intrahepatic vascular heterogeneity limits targeting efficiency; (2) Activation of compensatory pathways (e.g., FGF/Notch) leads to treatment resistance; and (3) Systemic administration carries off-target risks. Future breakthroughs hinge on developing biomarkers based on single-cell sequencing, applying novel models like the “hepatic hypertension-chip”, and implementing multi-target combination therapies.

Key Words: Liver cirrhosis; Portal hypertension; Pathological angiogenesis; Vascular endothelial growth factor; Hepatic stellate cells; Targeted therapy; GPR116; LECT2-Tie1 signaling pathway; Hepatic hypertension-chip

Core Tip: This review systematically elaborates the central role of pathological angiogenesis in cirrhotic portal hypertension, focusing on molecular mechanisms such as the hypoxia-inducible factor-vascular endothelial growth factor axis, inflammatory stress, and local renin-angiotensin system activation. Although targeted therapies show excellent efficacy in preclinical studies, their clinical translation is hampered by three challenges: Vascular heterogeneity, activation of compensatory pathways, and risks associated with systemic administration. The article also highlights emerging strategies, including targeting novel mechanosensory receptors such as GPR116, combination therapies, and precision medicine models based on the “hepatic hypertension-chip”, which offer new perspectives for individualized treatment.

INTRODUCTION

According to the latest data from the World Health Organization (WHO) for 2024-2025, approximately 307 million people worldwide are chronically infected with hepatitis B or C, which still causes about 1.3 million deaths annually from related end-stage liver diseases such as cirrhosis and liver cancer, indicating a persistently severe disease burden. Liver fibrosis and cirrhosis, stemming from viral hepatitis, alcohol abuse, and non-alcoholic fatty liver disease (NAFLD), among others, have become a major global health challenge. Patients with advanced cirrhosis often experience a severely diminished quality of life and increased mortality risk due to portal hypertension and its complications (e.g., liver failure, liver cancer)[1,2]. Current primary treatments (such as etiology management and liver transplantation) have limited efficacy, necessitating new strategies. In the progression of cirrhosis, pathological angiogenesis is a key driver of portal hypertension[3]. Abnormal neovessels accelerate disease progression by increasing intrahepatic resistance, promoting portosystemic shunting, and forming a vicious cycle of “angiogenesis-fibrosis-hypoxia”. Targeting dysregulated angiogenic pathways [e.g., the hypoxia-inducible factor-vascular endothelial growth factor (HIF-VEGF) axis, inflammatory-oxidative stress cascade, local renin-angiotensin system (RAS) activation] offers new hope for delaying or even reversing cirrhosis. This review aims to systematically analyze the molecular mechanisms of pathological angiogenesis in liver cirrhosis, evaluate the efficacy of current targeted therapeutic strategies, and discuss the challenges and future directions in their clinical translation.

LITERATURE SEARCH STRATEGY

To comprehensively and systematically review advances in the molecular mechanisms and targeted therapy of pathological angiogenesis in liver cirrhosis, a strict and reproducible literature search protocol was formulated. Searches were conducted across three authoritative databases: PubMed, Web of Science Core Collection, and the China National Knowledge Infrastructure (CNKI), to ensure coverage of international frontiers and significant domestic research. The search timeframe spanned from January 1, 2000, to November 30, 2025, to encompass key foundational studies and the latest developments in the field.

The search strategy employed a combination of controlled vocabulary (e.g., MeSH terms) and free-text terms to optimize sensitivity and specificity. The search strategy for PubMed was as follows: (“Liver Cirrhosis”[MeSH] OR “Hepatic Fibrosis”[MeSH] OR “liver cirrh*” [Title/Abstract]) AND (“Neovascularization, Pathologic”[MeSH] OR “pathological angiogen*” [Title/Abstract] OR “therapeutic angiogen*” [Title/Abstract]) AND (“Vascular Endothelial Growth Factor A” [MeSH] OR “VEGF” [Title/Abstract] OR “HIF-1alpha”[MeSH]) AND (“Molecular Targeted Therapy”[MeSH] OR “targeted therapy” [Title/Abstract]). An analogous strategy using corresponding Chinese keywords was applied to the CNKI database.

The literature inclusion criteria were: (1) Study subjects involving liver cirrhosis or related animal models; (2) Research content focusing on the molecular mechanisms of pathological angiogenesis or targeted therapeutic strategies (including pharmacological, gene, or cell therapies); and (3) Original research articles or systematic reviews published in Chinese or English. Exclusion criteria included: (1) Literature not in Chinese or English; (2) Publication types such as conference abstracts, case reports, commentaries, and articles with incomplete data; and (3) Literature irrelevant to the research topic or deemed low quality.

After initial retrieval using the above strategy, duplicate records were removed using reference management software (e.g., EndNote). Preliminary screening was performed by reviewing titles and abstracts. Subsequently, full-text articles of potentially eligible records were obtained and assessed for final eligibility during a secondary screening stage. Following this screening process, a total of 53 articles were ultimately included for analysis in this review. This rigorous process ensured the comprehensiveness, timeliness, and reliability of the content synthesized.

Core driver: The HIF-VEGF axis and its regulatory network

Chronic hypoxia is a core factor initiating pathological angiogenesis[4-6]. Liver tissue damage leads to the stable expression of HIF-1α, which acts as a “master switch” and directly upregulates the transcription of key genes, such as VEGF, thereby activating the entire angiogenic program[7,8]. After VEGF binds to its receptor VEGFR2, it strongly stimulates the proliferation, migration, and survival of liver sinusoidal endothelial cells (LSECs) through signaling pathways such as PI3K/Akt/mTOR[9], making it the most direct stimulus for angiogenesis[7,10]. The pathological significance of this axis lies in the “sinusoidal capillaryization” it triggers-the disappearance of LSEC fenestrations and the abnormal thickening of the basement membrane. This process significantly increases intrahepatic vascular resistance, forming the key structural basis for portal hypertension[11]. Notably, the HIF-VEGF axis does not operate in isolation; it is both an effector of hypoxia and a node integrating various signals, such as inflammation[5]. For example, studies found that neuropilin-1 can act as a co-receptor for VEGFR2[12], further amplifying VEGF signaling through the FAK/PI3K/Akt pathway and promoting abnormal vascular regeneration in liver fibrosis[13]. This suggests that targeting this axis may yield multi-faceted benefits but also hints that compensatory mechanisms may lead to treatment resistance[14].

Engine of a vicious cycle: The positive feedback between inflammatory-oxidative stress and angiogenesis

Inflammation and oxidative stress together constitute a self-amplifying vicious cycle, acting as a key engine driving pathological angiogenesis. Inflammatory cytokines such as tumor necrosis factor-alpha and interleukin-6 activate the nuclear factor kappa-B (NF-κB) pathway, not only directly damaging the endothelium but also further inducing VEGF expression, forming an “inflammation-angiogenesis” loop[13,15]. In this process, reactive oxygen species (ROS) derived from nicotinamide adenine dinucleotide phosphate (NADPH) oxidases (e.g., Nox2, Nox4) play an “accelerator” role[6,16,17]. ROS can not only directly activate NF-κB but are also induced by inflammatory cytokines themselves, thus forming a critical positive feedback loop: Inflammation → ROS production → NF-κB activation → release of more inflammatory cytokines/VEGF → further amplification of inflammation and oxidative stress. Additionally, oxidative stress can stabilize HIF-1α, synergizing with NF-κB to enhance VEGF signaling[10,13]. Ultimately, this vicious cycle leads to sustained abnormal vascular proliferation, and the newly formed leaky vessels exacerbate tissue hypoxia and inflammatory cell infiltration, perpetuating the cycle.

Other key pathways and vascular heterogeneity

Beyond the aforementioned central axis, a complex regulatory network finely controls angiogenesis, among which vascular heterogeneity is key to understanding therapeutic bottlenecks. LECT2-Tie1 pathway: A recent milestone study found that the secreted protein LECT2 is upregulated in cirrhosis and acts as a functional ligand for the orphan receptor Tie1. LECT2-Tie1 binding disrupts the Tie1/Tie2 heterodimer, activating downstream signals like MAPK, and differentially regulates vascular changes: It inhibits beneficial portal angiogenesis but promotes harmful sinusoidal capillarization[18]. This provides a new explanation for the limited efficacy of pan-VEGF inhibitors and suggests a new treatment strategy that requires distinct approaches for different vessel types. Notch and Wnt/β-catenin pathways: Notch signaling participates in pathological changes in sinusoids through mechanisms like regulating endothelial-mesenchymal transition[19,20]. The Wnt/β-catenin pathway primarily promotes hepatic stellate cell (HSC) activation, indirectly affecting the vascular environment[21,22]. These pathways engage extensively with the HIF-VEGF axis, collectively forming a complex regulatory network.

Central nexus: The RAS

Aberrant activation of the local hepatic RAS is another important aspect. Angiotensin II (Ang II) stimulates HSC activation and endothelial cell dysfunction via its AT1R. Ang II/AT1R signaling not only induces massive ROS production by NADPH oxidase[23,24], converging with the inflammatory-oxidative stress cycle, but also potently upregulates key pro-fibrotic factors like transforming growth factor beta 1[25], tightly coupling angiogenesis with the fibrotic process[26-28]. Given its central role, intervening in the RAS with ACEIs/ARBs has become an important direction in anti-fibrosis research[29-31].

Novel mechanisms and cutting-edge discoveries: Paradigm shift from “anti-angiogenesis” to “precision vascular regulation”

The application of single-cell RNA sequencing (scRNA-seq) has fundamentally shifted our understanding of liver pathophysiology, moving from a static catalog of cell types to a dynamic deconstruction of spatially and functionally heterogeneous ecosystems that drive disease progression. This is epitomized by the concept of the “pathological niche” sustained by imperfect repair, as seen in aging and fibrotic livers[32]. Applying this lens to cirrhosis, scRNA-seq studies have begun to map the precise cellular architecture of this niche. For instance, foundational work by Pietilä et al[4] in mice has revealed a previously unrecognized diversity of hepatic mesenchymal cells, with distinct subpopulations of HSC and fibroblasts located in different anatomical regions that exhibit divergent responses to injury[32].

The critical translational leap comes from etiology-stratified human studies, which have identified specific endothelial cell subpopulations whose molecular signatures are directly linked to clinical outcomes. A landmark study by Zhang et al[33] in hepatocellular carcinoma (HCC) demonstrated that within the vascular compartment of metabolic dysfunction-associated steatotic liver disease (MASLD)-related HCC, a subpopulation of tumor endothelial cells high in FABP4 expression was associated with vascular normalization, enhanced CD8+ T-cell infiltration, and a better response to immunotherapy[4].

The pivotal insight synthesizing these findings is that vascular heterogeneity is not merely a descriptive feature but a primary determinant of therapeutic efficacy and disease trajectory. The variable expression of key pathway components (e.g., VEGF receptors, angiopoietin-2, or etiology-specific markers such as FABP4) across these endothelial subpopulations provides a mechanistic explanation for the failure of one-size-fits-all anti-angiogenic therapies[33]. A therapy targeting a pathway active in one subpopulation (e.g., pro-fibrotic CD31hiPLVAPhiLSECs) may be ineffective or even detrimental to another subpopulation with regenerative potential (e.g., CLEC4GhiLSECs) or those responsible for immune cell recruitment. Therefore, the goal of scRNA-seq in cirrhosis is no longer just to identify heterogeneity, but to build an “etiology-stratified molecular atlas” that can classify patients based on the dominant pathobiological pathways active in their specific vascular niche, paving the way for truly precision medicine.

Recent technological advances, particularly single-cell sequencing, have fundamentally reshaped our understanding of vascular heterogeneity in cirrhosis, compelling a strategic pivot away from the blunt instrument of pan-anti-angiogenesis towards a nuanced paradigm of precision vascular regulation. This paradigm shift is powerfully illustrated by two landmark studies that offer complementary solutions to the challenge.

The study by Long et al[34], utilizing the innovative “hepatic hypertension-chip” (HH-Chip), represents a methodological breakthrough by directly interrogating the role of biomechanical forces. Their identification of GPR116 as a key mechanosensor in LSECs does more than just add a new molecule to the list of targets; it uncovers a primary driver of vascular injury-elevated hydrostatic pressure-that operates upstream of canonical growth factor signaling. The profound implication is that therapeutic strategies must evolve beyond merely blocking angiogenic factors to actively counteracting the physical forces that induce vascular dedifferentiation and regression. While the subsequent validation of GPR116 targeting via gene therapy[35,36] in animal models is promising, the translational hurdles are significant. These include developing safe and effective modalities to modulate GPR116 signaling in humans and ensuring that such interventions do not disrupt mechanosensing in other vascular beds.

Complementing this, the work by Xu et al[18] provides the molecular roadmap for executing precision regulation. By elucidating the LECT2-Tie1 axis, their research delivers a critical conceptual advance: Not all angiogenesis in cirrhosis is deleterious. The key pathological insight is that this pathway differentially regulates sinusoidal capillarization (harmful) and portal angiogenesis (potentially beneficial). This discovery effectively demolishes the monolithic view of angiogenesis and provides a scientific basis for therapies designed to selectively inhibit pathological vessels while sparing or even promoting restorative ones. However, the therapeutic application of this knowledge faces its own set of challenges, primarily the immense difficulty of designing drugs that can precisely disrupt the specific LECT2-Tie1 interaction without affecting the closely related and beneficial Tie2 signaling pathway.

In concert, these studies mark a decisive entry into a new era for cirrhotic vascular therapy. The HH-Chip/GPR116 axis answers the “how” of pathogenesis by revealing the biomechanical trigger of vascular decline. In contrast, the LECT2-Tie1 axis answers the “what” of pathological specificity by providing a criterion to distinguish friend from foe in the vascular landscape. The future frontier lies at the intersection of these approaches: Integrating mechanobiology with molecular subtyping to develop combination or sequential therapies that both protect LSECs from mechanical stress and precisely correct the aberrant signaling that drives pathological vessel formation. Overcoming the challenges inherent in targeting these novel pathways will be paramount to translating this revolutionary paradigm into clinical reality.

OVERVIEW OF MAJOR THERAPEUTIC STRATEGIES: CLASSIFICATION AND CHARACTERISTICS

Various targeted therapeutic strategies have been developed and are under investigation for pathological angiogenesis in liver cirrhosis. The table below systematically summarizes the core features of major strategies and their representative drugs/methods.

CHALLENGES AND FUTURE PERSPECTIVES

Although targeting pathological angiogenesis holds broad prospects for the treatment of liver cirrhosis, its clinical translation still faces severe challenges. Heterogeneity in treatment efficacy is a core issue, with therapeutic outcomes highly dependent on the disease stage. Additionally, compensatory drug resistance and the safety risks associated with systemic administration are significant obstacles[37-39]. It is particularly noteworthy that even with the discovery of novel targets through cutting-edge technologies like single-cell sequencing, the path to translation remains fraught with difficulties.

The translational gap from discovery to therapy: Deep-seated challenges

Firstly, inherent technological limitations pose a significant hurdle. Current research relies heavily on liver tissue samples from end-stage patients, which do not reflect the dynamic evolution of the disease at early stages and are subject to sampling bias. More critically, distinguishing the genuine driver pathogenic cell subpopulations that critically propel disease progression from merely stressed or reactive state cells within complex sequencing data necessitates more advanced bioinformatics algorithms and rigorous functional validation (e.g., using human organoids or other models for gene editing).

Secondly, the translation of biomarkers represents a core bottleneck. Histology-based “molecular atlases” are unsuitable for routine clinical monitoring. A pivotal next step is to convert tissue-based discoveries into non-invasive or minimally invasive liquid-biopsy biomarkers (e.g., endothelial cell-specific exosomes). However, this conversion is extremely difficult due to high background noise and low-abundance signals in the context of liver disease.

Finally, and most crucially, is the challenge of target “druggability”. Even after successfully identifying key targets such as the GPR116 or LECT2-Tie1 axis, developing agents that can precisely modulate their function-especially at protein-protein interaction interfaces-without disrupting related physiological pathways (e.g., Tie2 signaling) is far more challenging than developing traditional kinase inhibitors.

Therefore, future breakthroughs depend not only on mapping more perfect atlases but also on the synchronous innovation of targeted delivery technologies (e.g., liver-targeted nanocarriers) and the development of novel therapeutic modalities for “undruggable” targets (e.g., PROTAC technologies), to bridge the vast gap from “target identification” to “precision therapy”.

Future pathways for achieving breakthroughs

To break through the bottlenecks mentioned above, future research needs to focus on the following directions.

Leveraging cutting-edge technologies for deep patient stratification: Utilizing single-cell sequencing and artificial intelligence technologies to deeply analyze patient vascular heterogeneity and discover biomarkers predictive of treatment response (e.g., serum Ang II level)[40], thereby laying the foundation for individualized precision medicine[40-42].

Developing multi-target combination intervention strategies: Since inhibiting the main VEGF pathway often leads to feedback upregulation of alternative pathways, single-agent therapy has limited efficacy[43]. Therefore, combining anti-angiogenic drugs with anti-fibrotic drugs (e.g., pirfenidone) or anti-inflammatory drugs[44,45] shows greater promise than monotherapy, as it synergistically targets multiple pathogenic processes, including angiogenesis, fibrosis, and inflammation.

Innovating model systems and precision therapy platforms: Applying novel models such as the HH-Chip[34] to highly simulate the complex mechanical microenvironment of cirrhosis in vitro, providing a powerful platform for high-throughput drug screening and mechanistic research. Simultaneously, advancing frontier technologies, such as liver-targeted gene-editing systems[46], with a dual design of “liver-targeted delivery” and “hepatocyte-specific editing[47]”, offer new possibilities for fundamentally correcting the pathological process[32,48,49].

In summary, through the integration of precise delivery, combination therapies, novel target intervention, and frontier technologies, therapeutic strategies for pathological angiogenesis in liver cirrhosis are moving towards a more accurate and efficient new era, bringing hope for the eventual effective reversal of liver cirrhosis (Table 1)[50-53].

Table 1 Major therapeutic strategies for targeting pathological angiogenesis in liver cirrhosis.

Therapeutic strategy category	Representative drug/method	Primary target/mechanism	Development stage and evidence level	Key findings and existing challenges
Anti-VEGF therapy	Bevacizumab	Neutralizes circulating VEGF, blocking the VEGFR2 signaling pathway	Preclinical studies (e.g., CCl4-induced rat models): Reduces liver VEGF expression and portal pressure[28]. Phase II clinical trial: Showed reduction in hepatic venous pressure gradient, but with potential adverse effects like hypertension and proteinuria[29]	Findings: Effective in reducing portal pressure in animal models. Challenges: Clinical application is limited by systemic adverse effects (e.g., hypertension), particularly in portal hypertension patients, requiring caution
Multi-kinase inhibitor	Sorafenib	Inhibits multiple targets including VEGFR, PDGFR, Raf	Phase III clinical trials (approved for HCC). Trial in Child-Pugh B cirrhosis patients (REVERT): Did not significantly improve survival[50,51]	Findings: Improves sinusoidal capillarization in animal models. Challenges: Limited efficacy in advanced cirrhosis patients, indicating disease-stage dependent effectiveness
Targeting novel mechanisms	Targeting GPR116 (e.g., shRNA)	Inhibits the mechanosensor GPR116, protecting vascular integrity	Preclinical research (HH-Chip model and animal experiments)[34]	Findings: In the HH-Chip and animal models, inhibiting GPR116 alleviates pressure-induced LSEC injury and fibrosis, representing a shift from “anti-angiogenesis” to “anti-vascular regression”. Challenges: Early stage
Traditional Chinese medicine formula	Roucongrong granule (or equivalent)	Multi-target: Modulates LSEC lipid metabolism (CD36/PPAR pathway), inhibits HSC activation[52]	Preclinical research (animal models)	Findings: In CCl4-induced mouse cirrhosis model, improves liver microcirculation, inhibits HSC activation, downregulates VEGF/CD34 expression, showing multi-target synergistic effects. Challenges: Precise components/mechanisms and clinical trial evidence needed
Combination therapy	Anti-VEGF drug + anti-inflammatory drug (e.g., TNF-α inhibitor)	Simultaneously blocks angiogenesis and inflammation pathways	Preclinical research	Findings: Shows synergistic effects superior to monotherapy in animal models, more effectively reducing portal pressure. Challenges: Optimal combinations/timing and potential added toxicity need exploration
Interventional local delivery	Local VEGFR2 inhibitor post-TIPS	Local high-concentration delivery, reducing systemic toxicity	Exploratory research	Findings: Aims to reduce endothelial hyperplasia in TIPS shunt tracts, lowering restenosis risk[20,53]. Challenges: Delivery technology, carrier selection, and local safety are key issues

VEGF: Vascular endothelial growth factor; TNF-α: Tumor necrosis factor-alpha; VEGFR2: Vascular endothelial growth factor receptor 2; PDGFR: Platelet-derived growth factor receptor; Raf: Rapidly accelerated fibrosarcoma (kinase); HCC: Hepatocellular carcinoma; TIPS: Transjugular intrahepatic portosystemic shunt; LSEC: Liver sinusoidal endothelial cell; HSC: Hepatic stellate cell.

CONCLUSION

In conclusion, while the molecular landscape of pathological angiogenesis in cirrhosis is increasingly mapped, the journey from mechanistic insight to therapeutic breakthrough requires a paradigm shift in our approach. Success will depend on embracing patient stratification based on vascular molecular subtyping, investing in human-relevant disease models, and developing innovative delivery systems that maximize on-target effects while minimizing systemic collateral damage. The era of non-selective anti-angiogenesis is closing; the future lies in precision vascular medicine.

ACKNOWLEDGEMENTS

We extend our thanks to the peer scholars who provided valuable suggestions during the manuscript preparation, as well as the reviewers whose insightful comments significantly enhanced the quality of this paper.