Current evidence
Evidence for the role of GLP-1RAs in MASH-related HCC is sparse. Mouse models with biopsy-confirmed MASH and advanced fibrosis have demonstrated that semaglutide can reduce the incidence of HCC (40% vs. 88%) and tumor burden.23 Histological analysis revealed a reduction in Ki67-positive hepatocytes, suggestive of decreased cell proliferation, supported by a decrease in alpha-fetoprotein levels, a canonical biomarker of HCC. Notably, semaglutide did not induce any reversal in the fibrosis stage. These anti-tumor effects are likely mediated by systemic metabolic improvements rather than direct tumor targeting, given the expression patterns of GLP-1 receptor in tumor versus non-tumor tissues.
These results are consistent with earlier murine work by Kojima et al.,24 in which liraglutide completely abolished hepatocarcinogenesis in mice with streptozotocin- and high-fat diet-induced diabetes and MASH, whereas all control mice developed HCC.
Observational data in patients undergoing bariatric surgery further support these findings.25 Subsequent rates of both MASH and MASH-related HCC are significantly reduced, confirming the pivotal role of maintaining a healthy weight in preventing hepatocarcinogenesis. However, the paucity of clinical data on this phenomenon highlights an urgent need for clinical trials investigating the potential benefits of NuSHs in MASH and MASH-HCC by abolishing or reversing the oncogenic driver of this disease. If these agents were to find an adjunctive role in the oncological management of MASH-HCC, hard endpoints such as overall survival and progression-free survival would be warranted when comparing the addition of NuSHs against the current standard of care.
Proposed mechanism for the reversal of hepatocarcinogenesis
NuSHs represent particularly attractive candidates due to their multimodal effects on the metabolic derangements driving this disease continuum. By promoting weight loss, improving insulin sensitivity, and reducing systemic inflammation, these agents address key upstream drivers of oncogenesis in MASLD. Crucially, their ability to reduce hepatocyte stress and injury may create a microenvironment more conducive to immune surveillance and tumor suppression, potentially enhancing endogenous mechanisms of cancer control.
MASLD progresses through a chronic cycle of immune-mediated injury, in which hepatic inflammation, oxidative damage, and compensatory regeneration establish a permissive environment for malignant transformation. Oxidative stress induces mutagenic damage, while the regenerative response to chronic hepatocellular injury activates mitogenic pathways.26 Concurrently, immune tolerance to emerging neo-antigens establishes a tolerogenic microenvironment that allows transformed cells to escape immune surveillance. This immune–metabolic axis favors hepatocarcinogenesis.
However, intrinsic failsafe mechanisms are present to prevent malignant progression. Cell-intrinsic checkpoints (e.g., senescence, autophagy, apoptosis) are triggered in damaged hepatocytes. These processes release damage-associated molecular patterns (DAMPs) and senescence-associated secretory phenotypes (SASPs), recruiting and activating immune cells to clear damaged cells and reinforce tissue homeostasis.6 However, immune activity contributes directly to hepatic oxidative stress through reactive oxygen species, leading to DNA damage and further damage-associated molecular pattern release. This persistent inflammatory state increases the risk of oncogenic escape via accumulated secondary mutations.
Among the mediators of this immunometabolic loop, cytokines IL-6 and TNF-α are key players. Both are elevated in cirrhosis and HCC, where they activate STAT3 and NF-κB, transcription factors that mediate oncogenic inflammation.27 Kupffer cells, the liver’s resident macrophages, are a major source of IL-6, promoting hepatocyte proliferation, a process further supported by interleukin release from apoptotic hepatocytes (Fig. 2).
Single-cell RNA sequencing (scRNA-seq) studies in HBV-HCC have identified specific enriched tumor-associated macrophage (TAM) subsets (CCL18+ M2 macrophages) associated with disease progression, suggesting they help sculpt a tumor-permissive microenvironment.28 Terminally differentiated MMP9+ TAMs are also implicated.29 Further studies are needed to assess the role of these macrophages in MASH-HCC.
The proposed mechanisms by which GLP-1RAs and other NuSHs may reverse hepatocarcinogenesis remain speculative, as direct hepatic mechanisms are unclear. Canonical GLP-1R expression has not been consistently detected in hepatocytes, Kupffer cells, or hepatic stellate cells.30 This suggests that observed hepatic benefits may occur via indirect or extrahepatic mechanisms. Recent evidence indicates that GLP-1 may act via a gut-brain-immune axis, in which neuronal GLP-1Rs modulate systemic inflammation through autonomic pathways. Immune cells themselves express minimal GLP-1Rs, pointing to central modulation as a key anti-inflammatory mechanism.31
The downstream effects of GLP-1RAs on hepatic immune cells are not fully delineated. However, GLP-1RAs have been demonstrated to inhibit hepatic stellate cell activation by downregulating TGF-β1/SMAD and p38 MAPK signaling, thereby reducing fibrosis.32 Similarly, they can suppress NLRP3 inflammasome activation in Kupffer cells, leading to decreased IL-1β, IL-12, and TNF-α, and increased IL-10.33 This suppression of a pro-inflammatory state may disrupt the chronic inflammation-regeneration cycle that drives hepatocarcinogenesis.
Additionally, the therapeutic potential of GLP-1RAs and other incretin-based NuSHs in reversing the drivers of hepatocarcinogenesis in MASLD may be grounded in their ability to restore hepatic insulin sensitivity and correct metabolic dysregulation. Insulin regulates hepatic glucose and lipid metabolism through both direct and indirect mechanisms. Canonically, direct hepatic insulin action activates the insulin receptor tyrosine kinase, triggering PDK1 and mTORC2 signaling and culminating in AKT2 phosphorylation.34 This cascade promotes glucokinase translocation, glycogen synthase activation, and FOXO1 nuclear exclusion, thereby suppressing gluconeogenic gene expression. A refined model of insulin action further delineates that insulin also exerts indirect effects on gluconeogenesis via insulin receptor tyrosine kinase/AKT2 signaling in white adipose tissue, which suppresses lipolysis and reduces fatty acid and glycerol flux to the liver, thereby modulating gluconeogenesis through substrate availability and allosteric control.35 In MASLD, insulin resistance in both liver and adipose tissue disrupts these pathways, exacerbating steatosis and inflammation. GLP-1RAs, by enhancing insulin secretion and sensitivity, may reduce adipose lipolysis, restore hepatic AKT2 signaling, and suppress hepatic glucose production. In doing so, they target upstream metabolic drivers of inflammation, fibrosis, and hepatocarcinogenesis.
We propose that the use of NuSHs at the intermediate stage of MASH-HCC, following locoregional treatments, such as transarterial chemoembolization or selective internal radiotherapy, may delay commencing systemic anti-cancer therapy or even slow disease progression by removing the carcinogenic driver. There is also potential for increasing the number of patients who convert from unresectable to surgically resectable.
Proposed immunotherapy synergy as a mechanism of action
In patients with established MASH-related HCC requiring systemic anti-cancer therapy, NuSHs may still confer benefit. While their direct antitumor effects remain speculative and mechanistically unclear, the indirect benefits are compelling. These may include delays in progression to cirrhosis and liver decompensation—events that frequently necessitate pauses or cessation of systemic anti-cancer therapies, such as first-line palliative atezolizumab with bevacizumab—and improvements in metabolic parameters that may enhance tolerability of oncological treatments, particularly PD-1/PD-L1 immune checkpoint blockade.
A distinct CD8+CXCR6+ T cell subset has been shown to be highly enriched in the livers of patients with severe MASLD, particularly within MASH-HCC lesions, expressing elevated levels of PD-1, indicative of chronic activation and exhaustion.36 Strikingly, in preclinical MASLD models, PD-1/PD-L1 immune checkpoint blockade, while intended to restore antitumor immunity, paradoxically led to exacerbated liver injury, increased regenerative signaling, and a rise in tumor number and size, without meaningful tumor regression.36 These findings suggest that lipotoxicity and chronic hepatic inflammation characteristic of MASLD create an immune contexture that converts an otherwise favorable “hot” tumor microenvironment into a maladaptive one, promoting tissue damage and carcinogenesis rather than immune-mediated tumor clearance.37 This underscores the need for metabolic reprogramming alongside immunotherapy. GLP-1RAs and other incretin-based NuSHs, by ameliorating hepatic lipotoxicity, improving insulin sensitivity, and reducing inflammation, may help restore a more functional immune milieu. In this context, they hold promise not only for halting MASH progression but also for reconditioning the hepatic tumor microenvironment to enhance responsiveness to immunotherapies in MASH-HCC. However, both preclinical and patient-based assessments of their effects in this context are urgently needed.
Furthermore, improving metabolic status with NuSHs may reduce comorbidities such as diabetes, hypertension, and hypercholesterolaemia, as well as the risk of myocardial infarction and stroke. This could improve overall fitness, increasing the likelihood of eligibility for liver transplantation and thereby potentially curing patients of their MASH-HCC.
A summary of the established and potential roles of NuSHs in this setting is provided in Figure 3.