Hepatocellular carcinoma and vitamin D metabolism: novel targets and therapeutic strategies

Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related death worldwide. While age-standardized incidence and mortality have declined in some regions, the overall global burden continues to increase because of population aging and persistent etiologic factors. Curative options are limited to selected patients, and systemic therapies provide modest long-term benefit. Beyond its canonical role in calcium-phosphate homeostasis, vitamin D signals through the nuclear vitamin D receptor (VDR) to modulate immunity, oxidative stress, fibrosis, and cellular metabolism. In HCC, this axis is frequently dysregulated, including downregulation of CYP2R1, reduced CYP27B1 activity, upregulation of CYP24A1, and VDR dysfunction, which together blunt the antitumor actions of vitamin D and are linked to inflammation, aberrant lipogenesis, and immune evasion. Here, we summarize mechanisms by which vitamin D impacts key oncogenic pathways in HCC, including PI3K/AKT/mTOR, IL-6/STAT3, NF-κB, and TGF-β/SMAD, and highlight downstream nodes such as SREBP-1 and TXNIP as potential therapeutic targets. We also discuss emerging strategies to restore vitamin D signaling, such as CYP24A1 inhibition, next-generation vitamin D analogs, and VDR-biased agonists, to facilitate clinical translation and drug development.

Graphical Abstract

Highlights

Vitamin D metabolism and signaling are frequently dysregulated in hepatocellular carcinoma (HCC), including CYP2R1 downregulation, CYP27B1 suppression, CYP24A1 overactivation, and VDR silencing.

Vitamin D exerts antitumor effects in HCC through anti-inflammatory, antioxidant, antifibrotic, and metabolic reprogramming pathways.

SREBP-1 and TXNIP are identified as novel vitamin D–related molecular targets with potential therapeutic value in HCC.

Vitamin D analogs, CYP24A1 inhibitors, and VDR-biased agonists represent promising strategies for overcoming resistance and optimizing HCC treatment.

Introduction

Primary liver cancer is one of the most common malignancies in China, ranking fourth in incidence and second in cancer-related mortality nationwide [1]. According to the National Cancer Center of China (2022), approximately 367,700 new cases and 316,500 deaths were reported, and hepatocellular carcinoma (HCC) accounts for more than 90% of primary liver cancers [2]. Over the past two decades, widespread hepatitis B vaccination and improved screening have reduced age-standardized incidence and mortality; nevertheless, the absolute numbers of cases and deaths continue to increase because of population aging [3–5]. Men have an approximately threefold higher incidence and mortality than women [3, 6], and the burden remains higher in rural than in urban areas [6, 7]. National serosurveys indicate that HBsAg prevalence declined from 9.72% in 1992 to 5.86% in 2020, largely due to neonatal vaccination [8]. Despite progress in screening and early diagnosis, most patients are still diagnosed at intermediate or advanced stages, when opportunities for curative interventions (resection, ablation, or liver transplantation) are often missed. Early-stage, non-metastatic HCC may be cured by surgical or locoregional therapies, whereas intermediate or advanced disease typically requires systemic therapies, including tyrosine kinase inhibitors, immune checkpoint inhibitors, and combination regimens. These treatments can prolong survival (Fig. 1), but overall outcomes remain poor [9–11].

Fig. 1.

Current therapeutic standards for hepatocellular carcinoma. The current standards of care for hepatocellular carcinoma (HCC) are guided by the Barcelona Clinic Liver Cancer (BCLC) staging system, which categorizes patients into pre-metastatic and metastatic disease. In the pre-metastatic stages, treatment is predominantly based on physical or radiological interventions, including surgical resection, liver transplantation, radiofrequency ablation, and transarterial chemoembolization. In contrast, patients with metastatic disease are primarily managed with systemic anti-cancer therapies, such as targeted agents and immune checkpoint inhibitors

Vitamin D is a pleiotropic secosteroid hormone that regulates diverse physiological processes. Nearly eight decades ago, Apperly first reported an inverse geographic association in North America between sunlight exposure and overall cancer incidence and mortality [12]. In 1980, Garland and Garland proposed the “UVB–vitamin D–cancer” hypothesis, initially linking low sunlight exposure to an increased risk of colorectal cancer [13]. In 1992, Hanchette and Schwartz extended this inverse association to the geographic distribution of prostate cancer, further supporting the relationship between sunlight exposure, vitamin D status, and cancer risk [14]. Since then, ecological and observational studies have examined this hypothesis across more than a dozen cancer types, including breast [15], gastric [16], lung [17], pancreatic [18], ovarian [19], esophageal [20], bladder [21], and liver cancer [22]. Collectively, these data suggest that vitamin D deficiency is associated with an increased risk of cancer initiation and progression, and adequate circulating vitamin D levels have been linked to lower cancer risk and better clinical outcomes in observational studies [23].

However, translating these observations into clinical benefit has been challenging. First, calcitriol and some vitamin D analogs frequently cause hypercalcemia at antitumor doses, which is the major dose-limiting toxicity [24]. Second, several randomized controlled trials have not shown a reduction in overall invasive cancer incidence or cancer mortality with vitamin D supplementation, suggesting limited and heterogeneous effects in unselected populations [25]. Third, accumulating evidence indicates that cancer cells can decrease intracellular calcitriol availability through multiple mechanisms, thereby attenuating vitamin D–mediated antitumor activity [26, 27]. Accordingly, delineating how vitamin D metabolism and signaling become dysregulated in cancer is critical for developing strategies that overcome these limitations.

Beyond its classical role in calcium–phosphate homeostasis, vitamin D acts primarily through the nuclear vitamin D receptor (VDR), which is widely expressed across many cell types and regulates cellular differentiation, immune responses, redox balance, and inflammation [27–29]. Cellular homeostasis depends on coordinated vitamin D synthesis, activation, and degradation; disruption of this axis can compromise barrier integrity, immune tolerance, and metabolic balance [30, 31]. Hepatic vitamin D homeostasis is influenced by precursor acquisition (cutaneous synthesis and dietary intake) and by hydroxylation, activation, and catabolism in the liver and kidney [27, 32]. Importantly, tumor cells may hijack vitamin D signaling to blunt its antiproliferative and immunomodulatory functions, facilitating survival, immune evasion, and disease progression [33, 34]. Therefore, enzymes involved in vitamin D metabolism are being explored as potential therapeutic targets for HCC, both locally and systemically [35, 36].

In this review, we summarize current local and systemic therapeutic approaches for HCC and integrate evidence on vitamin D–related regulation of oncogenic signaling and metabolic reprogramming, highlighting potential translational opportunities in HCC.

Homeostatic hepatic vitamin D metabolism

Vitamin D metabolism comprises absorption, transport, activation, and inactivation and is central to calcium–phosphate homeostasis and immune regulation (Fig. 2) [37]. Dietary vitamin D₂ and D₃ are absorbed in the small intestine via bile salt–mediated micelle formation, primarily in the duodenum and jejunum [38]. Within enterocytes, vitamin D is packaged with triglycerides and cholesterol into chylomicrons that enter the circulation via the lymphatic system [39]. In the bloodstream, lipoprotein lipase (LPL) hydrolyzes triglycerides in chylomicrons, and vitamin D is delivered to the liver via chylomicron remnants [40]. In hepatocytes, vitamin D undergoes 25-hydroxylation by microsomal and mitochondrial cytochrome P450 enzymes, predominantly CYP2R1, generating 25-hydroxyvitamin D [25(OH)D], the major circulating storage form [41]. 25(OH)D is further hydroxylated in the kidney (and in certain immune cells) by mitochondrial 1α-hydroxylase (CYP27B1) to yield the active hormone 1,25-dihydroxyvitamin D [1,25(OH)₂D] [42]. CYP27B1 is regulated by parathyroid hormone (PTH), serum calcium and phosphate levels, and fibroblast growth factor 23 (FGF23) [43, 44]. Active 1,25(OH)₂D binds the vitamin D receptor (VDR), which heterodimerizes with retinoid X receptor (RXR) and binds vitamin D response elements (VDREs) in target genes, thereby regulating intestinal calcium absorption, bone mineralization, immune modulation, and cell proliferation [32]. Vitamin D is inactivated mainly by 24-hydroxylase (CYP24A1), which converts 25(OH)D and 1,25(OH)₂D into inactive metabolites (e.g., calcitroic acid) that are ultimately excreted via bile or urine, thereby maintaining metabolic homeostasis [45].

Fig. 2.

Vitamin D absorption and metabolism. Vitamin D originates from two primary sources. The first is cutaneous synthesis, whereby 7-dehydrocholesterol in the skin undergoes photolysis under UVB irradiation (290–315 nm), yielding previtamin D₃ that spontaneously isomerizes to vitamin D₃. The second is dietary intake of vitamin D₂ and D₃, which are absorbed in the small intestine via incorporation into mixed micelles or through transmembrane transporters, taken up by enterocytes, and subsequently secreted into the circulation within chylomicrons via the lymphatic system. A fraction of vitamin D can also enter the circulation through high-density lipoprotein (HDL) pathways. Newly synthesized vitamin D₃ diffuses from the skin into the bloodstream where it binds to vitamin D–binding protein (VDBP). In contrast, dietary vitamin D₂/D₃ is transported to the liver within chylomicrons. The first hydroxylation occurs in the liver, primarily catalyzed by CYP2R1, producing 25-hydroxyvitamin D [25(OH)D₃]. The 25(OH)D–VDBP complex is then filtered and reabsorbed in the proximal renal tubules, where mitochondrial CYP27B1 catalyzes the second hydroxylation, generating the hormonally active metabolite 1α,25-dihydroxyvitamin D₃ [1α,25(OH)₂D₃]. Inactivation is mediated by CYP24A1. This activation–inactivation cycle is tightly regulated by parathyroid hormone (PTH, stimulatory), fibroblast growth factor 23 (FGF23, inhibitory), circulating calcium and phosphate levels, as well as feedback from active vitamin D itself. Metabolic products are ultimately excreted through bile or urine. Beyond the kidney, extrarenal tissues such as the placenta and prostate also express CYP27B1, enabling local synthesis of 1α,25(OH)₂D₃ for paracrine and autocrine functions. Ultimately, active vitamin D binds to the vitamin D receptor (VDR), which heterodimerizes with retinoid X receptor (RXR) to regulate transcription of target genes involved in calcium–phosphate homeostasis, bone formation, immune modulation, and cell cycle control

Hepatic handling of vitamin D begins with 25-hydroxylation, in which CYP2R1 serves as the principal vitamin D 25-hydroxylase. Metabolic stressors such as obesity can suppress CYP2R1 transcription, thereby reducing circulating 25(OH)D concentrations [46]. Most circulating vitamin D metabolites are bound to vitamin D–binding protein (DBP), a liver-synthesized carrier that strongly influences transport and bioavailability [47]. Within the liver, VDR is enriched in non-parenchymal cells, including Kupffer cells, sinusoidal endothelial cells, cholangiocytes, and hepatic stellate cells, whereas quiescent hepatocytes express relatively low levels [48]. Bulk-liver analyses suggest that overall hepatic VDR expression increases in early metabolic-associated steatotic liver disease (MASLD) and declines as disease progresses toward metabolic-associated steatohepatitis (MASH) [49]. Because these measurements reflect mixed cell populations, the apparent changes may be driven by shifts in cellular composition and activation states, rather than uniform regulation of VDR within individual hepatic cell types.

Upon activation by 1,25(OH)₂D₃ or selective agonists, VDR promotes AMPK-dependent autophagy and restrains Akt–mTOR signaling, thereby reducing nuclear SREBP-1c activity and downregulating key lipogenic enzymes, including FASN, ACC, and ACLY, ultimately correcting dysregulated de novo lipogenesis in hepatocytes [50, 51]. VDR also modulates inflammatory and fibrogenic circuits by enhancing YAP1 activity to suppress NLRP3 signaling and by interacting with SMAD3 to antagonize TGF-β–SMAD3 signaling, thereby alleviating hepatic inflammation and fibrosis [52–54]. Consistently, VDR deficiency increases stellate cell activation and predisposition to hepatic fibrosis [55]. Taken together, hepatic CYP2R1-mediated 25-hydroxylation and cell type–specific VDR signaling integrate metabolic, inflammatory, and fibrotic programs that shape liver homeostasis and disease susceptibility.

Anticancer effects of vitamin D in liver cancer

Epidemiological and preclinical studies suggest that vitamin D may contribute to cancer prevention and treatment, prompting multiple mechanistic hypotheses to explain its antitumor effects. Accumulating evidence indicates that vitamin D can influence multiple stages of tumorigenesis, including initiation, progression, and metastasis, as well as dynamic interactions between tumor cells and the tumor microenvironment [56]. Mechanistically, vitamin D regulates intrinsic cellular programs—such as proliferation, differentiation, apoptosis, autophagy, and epithelial–mesenchymal transition (EMT)—and modulates extrinsic processes, including angiogenesis, oxidative stress responses, inflammation, and immune regulation [57]. In this section, we highlight representative pathways relevant to hepatocellular carcinoma (HCC) and summarize how vitamin D–VDR signaling may restrain tumor development and progression.

Anti-inflammation

In the HCC microenvironment, chronic inflammation is a key driver of tumor initiation and progression [58]. Among inflammatory circuits, the IL-6/STAT3 pathway plays a central role in propagating pro-tumorigenic signaling. Elevated IL-6 levels are associated with HCC onset, aggressive phenotypes, and poor prognosis [59]. Evidence indicates that the vitamin D–VDR axis can counteract this inflammatory milieu. In tumor-associated macrophages, vitamin D upregulates p27^Kip1, reduces intracellular IL-6 and TNF-α, suppresses downstream STAT3 activation, and ultimately attenuates tumor growth [60]. In addition, vitamin D has been reported to blunt IL-6- and TNF-α–mediated activation of TREM-1 and to downregulate NF-κB signaling in HCC cells, thereby reducing overall inflammatory activity [61].

Vitamin D can also induce THEM4, thereby inhibiting Akt phosphorylation and downstream IκBα phosphorylation in macrophages and suppressing NF-κB and COX-2 expression [62]. Notably, among hepatic non-parenchymal cells, macrophages exhibit the highest VDR expression, whereas hepatocytes express relatively low levels [63]. Consistently, VDR activation confers pronounced anti-inflammatory effects in murine hepatic macrophages, while VDR-deficient mice develop spontaneous hepatic inflammation. Collectively, these findings support a model in which macrophage-intrinsic VDR signaling restrains inflammation-driven, tumor-promoting signaling within the HCC microenvironment [63].

Antioxidant defense and DNA damage repair

Reactive oxygen species (ROS) contribute to hepatocarcinogenesis by promoting DNA damage and mutagenesis, supporting aberrant proliferation, triggering cell death, and amplifying inflammatory responses [64]. Accordingly, an effective antioxidant defense system is an important barrier to HCC initiation. Emerging evidence suggests that vitamin D enhances cellular antioxidant defenses, thereby mitigating oxidative stress and reducing the risk of DNA damage. Cathepsin L (CTSL), a lysosomal cysteine protease involved in protein degradation, antigen processing, and extracellular matrix remodeling [65], is upregulated in cancer and premature aging [66]. CTSL is expressed at higher levels in HCC tissues than in adjacent non-tumorous tissues, and increased CTSL expression correlates with reduced survival [67]. Mechanistically, vitamin D stabilizes the DNA repair factor 53BP1 and inhibits CTSL activity, thereby promoting repair of DNA double-strand breaks and providing support for its antitumor role in HCC [68].

Jia et al. further showed that vitamin D reduces SA-β-gal–positive cells, downregulates p53, p21, and p16, alleviates DNA damage, and promotes DNA repair via a Sirt1-dependent mechanism, consistent with suppression of cellular senescence [69]. Within the liver, VDR expression is enriched in non-parenchymal cells, particularly hepatic macrophages [63, 70]. In line with these observations, vitamin D attenuates bile duct ligation (BDL)–induced oxidative stress, hepatic fibrosis, and inflammatory cell infiltration [71]. Recent studies also indicate that the anti-apoptotic effect of vitamin D depends on autophagy and involves reduced ROS production through inhibition of the ERK/p38 MAPK pathway [72]. Moreover, vitamin D promotes macrophage autophagy by inhibiting the ROS–p38 MAPK axis, which suppresses inflammasome-mediated processing of IL-1β and alleviates inflammation [73]. Together, these data suggest that vitamin D supports protective programs during tumorigenesis by enhancing antioxidant responses, facilitating DNA repair, and coupling autophagy with inflammatory restraint.

Thioredoxin-interacting protein (TXNIP), originally identified as vitamin D–upregulated protein 1 (VDUP1), is transcriptionally induced by 1,25(OH)₂D₃ through direct activation of VDR [74, 75]. TXNIP contains an α-arrestin domain that interacts with cytosolic and mitochondrial thioredoxin (TRX), limiting TRX activity and thereby shaping redox signaling [76]. TXNIP also functions as a redox-sensitive regulator of proliferation and metabolism [77]. Because hepatocarcinogenesis is associated with hypermetabolism and oxidative stress [78], TXNIP downregulation in human HCC tissues and cell lines is of particular interest. Vitamin D₃ can induce TXNIP expression, suppress HCC cell proliferation, and potentially mitigate carcinogenic processes in chronic liver disease settings [79].

Recent work further demonstrates that 1,25(OH)₂D₃ promotes VDR binding to the TXNIP promoter and robustly induces TXNIP expression in cholangiocytes [80]. TXNIP upregulation suppresses proliferation markers (e.g., PCNA) and the anti-apoptotic protein Bcl-xL while promoting pro-apoptotic Bax, thereby shifting cells toward apoptosis. In parallel, TXNIP reduces TNF-α and TGF-β secretion, attenuating inflammatory and fibrogenic signaling to Kupffer cells and hepatic stellate cells (HSCs). Conversely, TXNIP loss in cholangiocytes increases TNF, IL-6, IL-1β, and chemokine production by Kupffer cells and enhances TGFB1 and collagen-related gene expression in HSCs, amplifying liver inflammation and fibrosis [80]. Collectively, these findings highlight the VDR–TXNIP axis as a mechanistically grounded target with potential therapeutic relevance in HCC.

Cell proliferation and metabolism

In HCC, the vitamin D–VDR axis is often described as a physiological “brake” on PI3K/Akt/mTOR signaling. Disruptions in vitamin D status and signaling—including deficiency, VDR downregulation, and abnormalities in metabolic enzymes—are frequently observed in HCC and can facilitate proliferation and metastasis [56, 81]. Class I PI3K converts PIP2 to PIP3 in response to receptor tyrosine kinase and insulin signaling, activating Akt and promoting mTOR complex activity [82]. Aberrant PI3K/Akt/mTOR signaling drives sterol regulatory element-binding protein 1 (SREBP-1) activation, which upregulates lipogenic enzymes such as SCD, FASN, ACC, and ACLY to support tumor growth [83]. Restoring vitamin D–VDR activity can counter this cascade by inducing PTEN and suppressing Akt phosphorylation [84]. In HCC cells, 1,25-dihydroxyvitamin D₃ increases PTEN expression via HDAC2 inhibition, thereby attenuating PI3K/Akt signaling and arresting cell-cycle progression, consistent with a VDR–PTEN–PI3K/Akt regulatory axis. Vitamin D can also resensitize everolimus-resistant HCC models, supporting functional crosstalk at the mTOR node and providing a rationale for combining vitamin D–VDR modulators with PI3K/mTOR inhibitors [36].

Vitamin D also intersects with hypoxia signaling. Calcitriol can modulate PTEN in a VDR-dependent manner, suppressing PI3K activity and indirectly limiting HIF-1/2α synthesis [85]. Notably, calcitriol also exerts VDR-independent effects: in HCC cells it reduces the translational efficiency of HIF-α mRNAs, redistributing HIF1A and EPAS1 transcripts to ribosomal fractions with lower translational activity, accompanied by reduced phosphorylation of Akt and downstream translation initiation factors [86, 87]. In addition, vitamin D induces DDIT4 (REDD1), promoting assembly of the TSC1/2 complex and suppressing mTOR activity [88]. Because HIF-1 can itself induce DDIT4/REDD1, this creates a negative feedback loop that restrains mTORC1 activity and limits HIF-1α accumulation, thereby constraining tumor growth [89]. Collectively, these findings indicate that vitamin D can restrain hypoxia-adaptive growth through both canonical VDR-dependent signaling and non-canonical mechanisms, including DDIT4/REDD1-mediated feedback. Overall, restoring vitamin D signaling may help suppress PI3K/Akt/mTOR-driven hepatocarcinogenesis and disease progression.

SREBP-1 (SREBF1) is a master transcription factor that drives de novo lipogenesis and exists mainly as the SREBP-1a and SREBP-1c isoforms [90]. In the endoplasmic reticulum, the precursor associates with SREBP cleavage–activating protein (SCAP). When cellular lipids are depleted, the SREBP–SCAP complex translocates to the Golgi, where sequential cleavage by S1P and S2P releases the nuclear-active SREBP-1 form, upregulating lipogenic genes (e.g., ACACA, FASN, SCD1) and supporting membrane synthesis, energy balance, and reducing equivalents [91]. Oncogenic pathways, including PI3K/Akt/mTOR, promote SREBP-1 activation, and in turn SREBP-1 reinforces a metabolically adaptive “lipogenic phenotype” that supports tumor growth [92]. Recent studies also implicate tumor-intrinsic networks (e.g., SOCS5–RBMX) in transcriptional activation of SREBF1 and highlight protein degradation pathways as additional regulatory layers, underscoring SREBP-1 as a plastic metabolic node and potential therapeutic vulnerability [93].

In the canonical cholesterol-sensing pathway, cholesterol binding to SCAP promotes SCAP–Insig interaction at the ER membrane, preventing SREBP–SCAP transport to the Golgi and thereby suppressing SREBP activation [94]. Asano et al. reported that 25-hydroxyvitamin D [25(OH)D] physically interacts with SCAP, triggers cleavage of its C-terminal fragment, and promotes polyubiquitination and proteasomal degradation [95]. Because SCAP is required for activation of all SREBP isoforms, SCAP degradation reduces both precursor and mature SREBP levels, suppresses transcription of downstream lipogenic targets, and decreases cellular pools of key fatty acids, lowering overall de novo lipogenesis [96]. In HCC, targeting the vitamin D–SCAP–SREBP-1 axis may disrupt SREBP-1–dependent lipogenic addiction and undermine tumor lipid supply at a metabolic root. These findings provide a rationale for optimizing vitamin D status and for developing 25(OH)D-based SCAP–SREBP inhibitors that may act independently of VDR [92, 95, 97].

Vitamin D may also modulate tumor vascular plasticity. Vitamin D–binding protein (VDBP) can interact with the helix–loop–helix DNA-binding domain of Twist1, blocking Twist1 binding to the VE-cadherin promoter and inhibiting vasculogenic mimicry (VM). Vitamin D also promotes formation of a VDR–YY1 complex that enhances transcription of the GC gene and increases VDBP expression [98]. This cascade can suppress tumor proliferation and limit vascular plasticity and immune exclusion. In patient-derived xenograft (PDX) models, vitamin D combined with anti–PD-1 therapy shows greater antitumor efficacy than immune checkpoint blockade alone [98]. These results highlight the therapeutic potential of the VDR–YY1–GC/VDBP axis and support combining vitamin D pathway activation with immunotherapy to improve outcomes in HCC.

Anti-fibrotic effects

The TGF-β/SMAD cascade is a prototypical pro-fibrotic pathway that is frequently hyperactivated in HCC, driving hepatic stellate cell activation, extracellular matrix deposition, EMT, and tumor progression [99, 100]. In this pathway, receptor-activated SMAD2/3 transduce TGF-β signals to chromatin with SMAD4, whereas SMAD7 acts as a negative regulator [101]. SMAD3 serves as a central transcriptional effector of hepatic fibrogenic programs [102]. From a hepatopathological perspective, dysregulated vitamin D metabolism may remove an endocrine “brake” on pro-fibrotic TGF-β/SMAD signaling, thereby predisposing the liver to fibrosis-driven hepatocarcinogenesis. In a hepatic fibrosis model, vitamin D/VDR directly bound VDREs within the HRC promoter, suppressed HRC transcription, attenuated TGF-β1/SMAD3 signaling, and inhibited stellate cell activation [103]. Conversely, HRC overexpression enhanced TGF-β1/SMAD3 signaling while downregulating SMAD7, accelerating S-phase entry and proliferation; related observations were also reported in models with enforced Smad7 expression [104].

To further dissect vitamin D–TGF-β interactions during hepatocarcinogenesis, Chen et al. showed that vitamin D deprivation aggravated tumor burden in an HCC mouse model with Smad3 deficiency, accompanied by upregulation of TLR7 [105]. Mechanistically, Smad3 bound SBE sites within the TLR7 promoter, and TGF-β stimulation strengthened this interaction and suppressed TLR7 promoter activity. Notably, vitamin D retained the ability to repress TLR7 signaling even under low Smad3 activity [105]. TLR7 silencing reduced HCC proliferation and migration, whereas combined vitamin D deficiency and Smad3 impairment synergistically activated Wnt/β-catenin signaling and increased oncogenic proteins such as Akt/mTOR, together fostering a pro-inflammatory, pro-fibrotic, tumor-promoting microenvironment [78, 105].

Vitamin D can also attenuate fibrosis by activating VDR in hepatic stellate cells (HSCs). VDR interferes with early YAP transcriptional activity, preventing HSC transdifferentiation into myofibroblasts, and activates AMPKα, reducing ATP production and limiting HSC proliferation [106]. In CCl₄-induced fibrosis, liver-specific deletion of YAP/TAZ reduces fibrosis and diminishes the protective effects of VDR activation, whereas YAP inhibition rescues the exacerbated fibrosis caused by VDR deficiency [106]. Combined treatment with a VDR agonist and a YAP inhibitor synergistically suppresses HSC activation and liver injury [106]. Furthermore, the non-steroidal VDR agonist 16i reduces fibroblast activation protein-α (FAPα) expression and disrupts the positive feedback loop between neutrophil extracellular traps (NETs) and HSC activation. This effect is mediated by a VDR–RXR complex that competitively limits transcriptional activator binding at the FAPα promoter, thereby suppressing fibrogenic signaling and inflammatory cytokine production [107]. Together, these studies suggest that restoring vitamin D–VDR activity may simultaneously dampen fibrosis-related and tumor-intrinsic pathways, offering opportunities to prevent or slow hepatocarcinogenesis (Fig. 3).

Fig. 3.

Regulatory roles of vitamin D in HCC signaling. Vitamin D exerts antitumor effects in hepatocellular carcinoma (HCC) through multiple regulatory mechanisms. The PI3K–Akt–mTOR pathway drives the activation of SREBP-1 and upregulates lipogenic enzymes such as ACLY, ACC, FASN, and SCD, thereby promoting cell proliferation, migration, and drug resistance. vitamin D counteracts this process by inducing the expression of PTEN, DDIT4, and HIF-α, leading to inhibition of Akt/mTORC1 signaling, and further constrains SREBP-1 activation by suppressing SCAP activity, effectively attenuating the “lipogenic” phenotype of tumor cells. In parallel, vitamin D negatively regulates the TGF-β/Smad signaling axis, which contributes to the attenuation of hepatic fibrosis and the suppression of tumor progression

Dysregulation of vitamin D activity in HCC cancer

Traditionally, CYP27B1 has been viewed as predominantly expressed in the kidney, where it catalyzes calcitriol synthesis, whereas VDR and CYP24A1—key mediators of calcitriol action and degradation—were thought to be largely restricted to classical target tissues (e.g., intestine, bone, and kidney). However, accumulating evidence indicates that these proteins are also expressed in non-renal, non-classical tissues, including hepatocytes and liver tumor tissues [108, 109]. Accordingly, hepatocytes and their malignant derivatives may represent direct targets of vitamin D signaling. Notably, local regulation of vitamin D activity is frequently disrupted in HCC cells, which can diminish responsiveness to vitamin D–based interventions and contribute to therapeutic resistance [36, 110]. In this section, we summarize current evidence for dysregulated vitamin D metabolism and signaling in HCC, as outlined in Fig. 4.

Fig. 4.

Disrupted vitamin D metabolism in HCC

Disruption of the VDR–RXRα axis undermines vitamin D signaling in HCC

To exert canonical transcriptional activity, the vitamin D receptor (VDR) must heterodimerize with retinoid X receptor-α (RXRα) to form a transcriptionally competent complex. This complex recognizes DR3-type vitamin D response elements (VDREs) and activates or represses downstream target genes [111]. Therefore, RXRα conformation and post-translational modifications are critical determinants of VDR signaling output. Aberrant RXRα phosphorylation has emerged as a central mechanism that disrupts VDR signaling in hepatocellular carcinoma. In clinical HCC tissues and cell models, RXRα is strongly phosphorylated at Ser260 by the ERK/MAPK pathway, which impedes its degradation, reduces transcriptional activity, and promotes tumor growth [112]. Mechanistically, RXRα-Ser260 phosphorylation compromises recruitment of co-activator complexes (e.g., DRIP205/MED1), destabilizes VDR–RXR–chromatin interactions, and induces a “hormone resistance-like” state against 1,25(OH)₂D₃ [113, 114]. Consistently, MAPK inhibition or expression of a non-phosphorylatable RXRα mutant (S260A) restores nuclear localization and co-activator assembly of this axis [113, 114]. In vivo, phosphorylated RXRα (p-RXRα) promotes DEN-induced hepatocarcinogenesis via activation of β-catenin signaling, supporting p-RXRα as a potential therapeutic target in HCC [115]. Notably, although somatic mutations in the RAS/RAF/MAPK cascade are relatively infrequent in HCC, pathway activation is observed in more than 50% of cases, providing an upstream context for aberrant RXRα phosphorylation [116].

In HCC, VDR expression shows marked cohort-dependent variability. Some studies report relatively elevated VDR mRNA expression in non-parenchymal cells within HCC tissues, suggesting that higher VDR abundance may increase cellular sensitivity to the antiproliferative effects of vitamin D [117]. However, evidence also indicates dysregulation of the KLF4–VDR axis during liver disease progression, with KLF4 downregulation impairing transcriptional control of VDR [118]. These findings highlight that VDR expression levels do not necessarily reflect functional vitamin D signaling, as transcriptional output is contingent on RXRα status and the integrity of associated co-regulators. In HBV-related HCC, HBV infection can downregulate VDR transcription and protein levels in hepatoma cells, thereby attenuating the inhibitory effects of vitamin D on viral transcription–replication and inflammatory signaling [119]. Conversely, activated VDR can directly suppress HBV core promoter activity, indicating bidirectional virus–host regulation at the level of VDR [120]. In addition, increased CpG methylation within the VDR promoter has been reported in HCC compared with chronic liver disease tissues; demethylating treatment reduced promoter methylation and restored VDR expression in cellular models, suggesting that epigenetic silencing is a potentially reversible mechanism of VDR inactivation in HCC [121].

Reduced CYP27B1-mediated vitamin D activation in HCC

CYP27B1 catalyzes conversion of 25(OH)D to the active metabolite 1,25(OH)₂D [122]. Although classically considered kidney-dominant, CYP27B1 is also expressed in multiple extra-renal tissues, enabling local vitamin D activation and paracrine/autocrine regulation within tissues and tumor microenvironments. In liver-related studies, Chiang et al. detected CYP27B1 in HCC cell lines (e.g., HepG2) and human HCC tissues, and demonstrated that 25(OH)D can be converted to 1,25(OH)₂D in HepG2 cells, resulting in induction of canonical target genes such as CYP24A1 [123]. Clinically, CYP27B1 mRNA levels are significantly lower in HCC tumor tissues than in non-tumorous counterparts, suggesting reduced local activation capacity with malignant progression [109].

The HCC microenvironment is characterized by chronic inflammation and cytokine-driven signaling, particularly persistent activation of the IL-6/STAT3 axis, which is closely linked to tumor initiation, progression, and therapeutic responses [59]. Although direct evidence for cytokine regulation of CYP27B1 in human HCC cells remains limited, studies in colorectal cancer and primary human placental cells show that TNF-α and IL-6 can downregulate CYP27B1 transcription, supporting the concept that inflammatory networks may attenuate local vitamin D activation [124, 125]. This mechanism is biologically plausible in inflammation-rich HCC but warrants validation in liver cancer models. Notably, immune cells such as macrophages and dendritic cells may upregulate CYP27B1 in response to inflammatory stimuli, in contrast to the suppressive effects observed in epithelial tumor cells, indicating cell type–specific regulatory patterns [126]. Such bidirectional regulation may collectively shape vitamin D metabolism within HCC.

CYP2R1 downregulation associated with HCC progression

CYP2R1 is the predominant hepatic vitamin D 25-hydroxylase, catalyzing conversion of cholecalciferol to 25(OH)D, the principal circulating precursor for VDR signaling [127]. Under metabolic stress conditions closely linked to hepatocarcinogenesis (e.g., obesity, diabetes, or fasting), hepatic CYP2R1 expression is suppressed. This suppression is associated with reduced circulating 25(OH)D levels and may attenuate the antiproliferative and antifibrotic effects of vitamin D [46, 128]. Epidemiological studies further show that lower serum 25(OH)D concentrations are associated with higher HCC incidence and adverse liver outcomes, whereas bioavailable vitamin D levels correlate strongly with survival in HCC patients, suggesting potential biomarker utility during progression from metabolic dysfunction–associated steatotic liver disease (MASLD) to HCC [129]. Genetic evidence also indicates that CYP2R1 polymorphisms modulate HCC risk in patients with HCV infection, implying that impaired 25-hydroxylation capacity may increase susceptibility to hepatocarcinogenesis in the setting of chronic liver injury [130]. Collectively, these findings suggest that restoring CYP2R1 activity or hepatic 25-hydroxylation capacity may represent a potential adjunctive strategy to limit HCC initiation and progression.

CYP24A1 upregulation and vitamin D resistance in HCC

CYP24A1 is a mitochondrial 24-hydroxylase that terminates vitamin D signaling by degrading 25(OH)D and 1,25(OH)₂D, thereby attenuating VDR-mediated antiproliferative effects [32]. In hepatocarcinogenesis, particularly under hypoxic conditions, calcitriol markedly induces CYP24A1 expression in HCC cells (e.g., Huh7), establishing a feedforward “vitamin D resistance” mechanism that compromises VDR activity [131]. A 2024 multi-cohort machine learning study assigned positive weights to CYP24A1 in an HCC risk-scoring model; higher scores correlated with poorer overall survival, supporting CYP24A1 as a potential prognostic biomarker for HCC progression from MASLD or viral liver injury [132]. Genetic studies further implicate the CYP24A1 rs6013897 polymorphism as being associated with cirrhosis and HCC risk, consistent with a role for altered vitamin D catabolism in tumorigenesis under chronic liver injury [133]. Angeli-Terzidou et al. reported that flavonoids suppress CYP24A1 expression, thereby restoring 1,25(OH)₂D₃ activity and anticancer effects in HCC cells [131]. Together, these findings position CYP24A1 as both a “brake” on vitamin D signaling and a potential oncogenic amplifier in HCC. Targeted inhibition of CYP24A1 may prolong the intracellular half-life of calcitriol and potentiate its tumor-suppressive activity [131, 134]. Accordingly, strategies combining vitamin D supplementation with CYP24A1 inhibitors may offer new avenues for overcoming resistance in HCC.

Therapeutic prospects of vitamin D in hepatocellular carcinoma

Although randomized clinical trial evidence for HCC is currently lacking, epidemiological data, clinical observations, and preclinical models consistently indicate that activation of vitamin D signaling may play a pivotal role in the prevention and treatment of HCC [17, 22, 135]. In recent years, multiple potential intervention strategies have been proposed targeting dysregulated vitamin D metabolism and functional impairment within the tumor microenvironment. However, vitamin D–based therapies in HCC remain constrained by important limitations, including dose-dependent toxicity, pharmacokinetic variability, and adaptive resistance of tumor cells to vitamin D signaling. Therefore, elucidating the molecular mechanisms underlying dysregulated vitamin D metabolism and signaling in HCC, and exploring combinatorial strategies with existing therapies, will be critical for advancing the clinical translation of vitamin D.

In HCC, systemic activation of the vitamin D–VDR signaling axis is similarly constrained by dose-limiting hypercalcemia, prompting a paradigm shift in therapeutic strategies from “potency enhancement” to “optimization of selectivity.” Two approaches are of particular significance: (i) developing VDR-biased agonists with minimal or no calcemic activity, and (ii) modulating metabolic pathways to increase the local availability of active metabolites within tumors. Within the first strategy, the classical non-calcemic analogue seocalcitol has advanced to phase II clinical trials in unresectable HCC. However, the overall objective response rate remained below the 20% efficacy threshold defined by the Gehan design, indicating limited therapeutic benefit. Hypercalcemia was the most common adverse event, affecting more than 80% of patients [136]. Although in vitro studies have demonstrated that vitamin D can exert antifibrotic effects by suppressing aberrant activation of the TGF-β1 signaling pathway, emerging clinical evidence has yielded inconsistent results. In contrast to findings from cell-based experiments, multiple clinical studies have shown that weekly supplementation with 60,000, 80,000, or 100,000 IU of vitamin D₂ for six consecutive weeks does not significantly reduce serum levels of TGF-β1 or other profibrotic markers compared with placebo [137, 138]. In addition, a single dose of vitamin D has been shown to significantly reduce circulating proinflammatory microRNAs; however, no change in serum miR-21 levels was observed compared to baseline [139]. This discrepancy suggests that the antifibrotic effects observed in vitro may be constrained in clinical settings by limited achievable dosing, impairment of downstream signaling pathways, and the complexity of the disease microenvironment. Moreover, the CYP11A1-driven non-classical vitamin D metabolic pathway has been considered a potential avenue for the development of biased agonists. Its polyhydroxylated metabolites, such as 20(OH)D₃ and 20,23(OH)₂D₃, have been reported in non-hepatic models to concurrently modulate VDR and RORα/γ signaling, exhibiting antiproliferative and prodifferentiation effects with minimal risk of hypercalcemia [140–142]. Although this discovery provides an important direction for optimizing low-calcemic ligands in HCC, direct evidence within hepatic cancer models and the relative contributions of VDR compared to ROR still require systematic clarification. In addition, vitamin D reverses acquired resistance to the mTOR inhibitor everolimus by upregulating miR-375 and downregulating resistance-associated pathways such as MTDH/YAP1/c-MYC, providing molecular rationale for its combination with targeted therapies [36]. Epidemiological evidence further shows that bioavailable 25(OH)D levels, rather than total concentrations, are independently associated with survival outcomes in HCC patients. This indicates that future clinical trials should incorporate stratification by vitamin D status and integrate metabolic enzyme and receptor characteristics (e.g., VDR expression, methylation status, and CYP24A1/CYP27B1 phenotypes) to more accurately identify responsive subgroups and refine efficacy–safety evaluations [121, 143].

As noted above, dysregulated vitamin D metabolism plays a critical role during HCC progression. Aberrant activation of the PI3K/Akt/mTOR pathway drives overexpression of lipogenic enzymes such as FASN, ACC, SCD, and ACLY, whereas vitamin D induces PTEN expression and suppresses Akt phosphorylation, thereby effectively blocking this cascade and restraining tumor metabolic reprogramming. Moreover, vitamin D regulates the translational efficiency of HIF-1/2α mRNA through VDR-independent mechanisms, thereby reducing phosphorylation of Akt and downstream translation initiation factors, further limiting tumor adaptive growth under hypoxic conditions. Within the TGF-β/SMAD pathway, vitamin D binds to VDRE sites in the HRC promoter to suppress TGF-β1/SMAD3 signaling and inhibit HRC transcriptional activity, thereby halting hepatic fibrogenesis and lowering HCC risk. Thus, therapeutic development targeting the VDR–PI3K/Akt/mTOR and VDR–TGF-β1/SMAD axes holds promise as a potential strategy for HCC treatment. Meanwhile, aberrant expression of key metabolic enzymes such as CYP2R1 and CYP24A1 further disrupts vitamin D homeostasis in HCC, suggesting that modulators or inhibitors targeting these enzymes may be valuable in restoring vitamin D metabolic balance. In addition, proteins such as SREBP-1 and TXNIP, which function as central nodes in vitamin D–mediated regulation of metabolism and oxidative stress, also show potential as novel therapeutic targets in HCC, providing new avenues for drug development. Current HCC management relies primarily on potentially curative approaches at early stages, such as surgical resection, liver transplantation, and radiotherapy. However, once the disease advances, existing interventions largely prolong survival without achieving cure. Management of HCC remains a major clinical challenge, yet providing effective therapeutic opportunities remains critical for a disease often regarded as “incurable.” Therefore, vitamin D–based therapeutic strategies are increasingly regarded as potentially valuable adjuncts in the management of hepatocellular carcinoma. Given its biological roles in immunomodulation and antifibrotic processes, further investigation is warranted. Nevertheless, more robust clinical evidence is required to validate their efficacy in improving patient outcomes and reducing disease-specific mortality.

Author contributions

H.G. conceptualized the review and drafted the main content. H.G., G.W., Z.C and L.H, conducted the literature search and analysis. G.W., H.G. contributed to the organization of the manuscript and provided significant input on specific sections. All authors contributed to the final revision and approved the submitted version.

Funding

This work has been supported by grants: Zhejiang Shuren University Basic Scientific Research Special Funds (KXJ1724104C and 2024XZ013), The opening foundation of the State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, The First Affiliated Hospital, Zhejiang University School of Medicine (SKLID2024KF07) and Talent Introduction Project of Zhejiang Shuren University (KXJ1723105).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.