From kidney to liver: Are sodium–glucose cotransporter-2 inhibitors emerging as metabolic disease modifiers?

Abstract

BACKGROUND

Sodium–glucose cotransporter-2 (SGLT2) inhibitors are increasingly recognized for benefits beyond glycaemic control, with established roles in renal and cardiovascular protection. In diabetic kidney disease (DKD), dapagliflozin-based strategies target persistent proteinuria, while emerging data suggest potential effects on hepatic outcomes, including hepatocellular carcinoma (HCC), in metabolic dysfunction–associated steatotic liver disease (MASLD).

AIM

To evaluate the association between SGLT2 inhibitor use and HCC incidence in MASLD.

METHODS

A systematic review and meta-analysis was conducted following PRISMA 2020 guidelines and registered in PROSPERO (CRD1003646). Studies including adults with MASLD comparing SGLT2 inhibitors with other oral hypoglycemic agents and reporting HCC incidence were identified through searches of PubMed, EMBASE, Scopus, and Web of Science up to April 12, 2025. Pooled risk ratios (RRs) were calculated using a random-effects model, and heterogeneity was assessed using the I² statistic.

RESULTS

Three retrospective cohort studies comprising 352890 individuals with MASLD were included. SGLT2 inhibitor use was associated with a significantly lower risk of HCC (RR = 0.60, 95%CI: 0.52-0.70; P < 0.0001), corresponding to an approximate 40% relative risk reduction, with low heterogeneity (I² = 14.1%). Sensitivity analyses yielded consistent results.

CONCLUSION

SGLT2 inhibitor use is associated with a reduced incidence of HCC in MASLD, although findings are derived from observational data. These results, together with established renal and cardiovascular benefits, support a broader role for SGLT2 inhibitors as modifiers of metabolic disease across organ systems.

Key Words: Hepatocellular carcinoma; Sodium-glucose transporter 2 inhibitors; Metabolic dysfunction-associated steatotic liver disease; Chronic kidney disease; Metabolic disease modification

Core Tip: Sodium-glucose cotransporter-2 inhibitors extend beyond glycaemic control and may modify key pathways of metabolic disease. Evidence from diabetic kidney disease and metabolic dysfunction-associated steatotic liver disease suggests these agents attenuate inflammation and fibrosis, with a potential association with reduced hepatocellular carcinoma incidence. This supports a shift toward integrated, cross-organ metabolic disease modification.

INTRODUCTION

We read with great interest the recent study by Pasari et al[1] which evaluated the additive effect of finerenone in patients with diabetic kidney disease (DKD) receiving maximally tolerated doses of dapagliflozin and telmisartan, addressing the persistent challenge of residual proteinuria despite optimized therapy. This work highlights a critical unmet need in metabolic disease: Even with the use of mature kidney protection therapies, the persistent inflammation and fibrotic activities will continue to cause organ damage. Beyond its immediate nephrological implications, this study prompts broader reflection on the evolving role of sodium-glucose cotransporter-2 (SGLT2) inhibitors, which have transitioned from glucose-lowering agents to central components of disease-modifying strategies targeting the upstream drivers of metabolic injury. In this context, we could not help but reflect on the expanding role of SGLT2 inhibition across multiple facets of metabolic disease, particularly in conditions where shared pathophysiological pathways link organ-specific complications.

Hepatocellular carcinoma (HCC) exemplifies one such complication, remaining the third leading cause of cancer-related mortality worldwide, with its epidemiology increasingly shaped by metabolic dysfunction-associated steatotic liver disease (MASLD)[2,3].The mechanisms underpinning MASLD-related hepatocarcinogenesis-insulin resistance, chronic inflammation, oxidative stress, and fibrogenesis-closely mirror those driving DKD progression, reflecting a common metabolic substrate of disease[4-8]. Despite the growing burden of MASLD-associated HCC, effective pharmacologic strategies for cancer prevention remain lacking, with current approaches largely limited to lifestyle interventions and indirect metabolic modulation[9-11]. In this setting, SGLT2 inhibitors have emerged as promising agents with pleiotropic effects, including improvements in steatosis, inflammation, and fibrosis, and accumulating observational data suggest a potential reduction in HCC incidence among treated patients[12-14]. While recent advances such as the approval of resmetirom and semaglutide for metabolic dysfunction–associated steatohepatitis (MASH) represent important progress[15,16], no therapy has yet been established for HCC prevention. Taken together, the findings of Pasari et al[1] and the emerging hepatology data support a unifying concept: SGLT2 inhibitors may act as systemic modifiers of metabolic disease, with the potential to attenuate downstream complications across organs, including both renal fibrosis and hepatic carcinogenesis. The aim of this systematic review is to evaluate the association between SGLT2 inhibitor use and HCC incidence in MASLD.

MATERIALS AND METHODS

To contextualize the emerging role of SGLT2 inhibitors as potential modifiers of metabolic disease across organ systems, we conducted a systematic review and meta-analysis focused on their association with HCC incidence in MASLD. The study was prospectively registered in the PROSPERO (CRD1003646) and followed the PRISMA 2020 framework. Eligible studies included adult populations (≥ 18 years) with MASLD in which SGLT2 inhibitor therapy was compared with alternative oral hypoglycemic agents, with HCC incidence defined as the primary outcome. A comprehensive literature search was performed across PubMed, EMBASE, Scopus, and Web of Science up to 12 April 2025: “Sodium-Glucose Transporter 2 Inhibitors”[MeSH] OR “SGLT2 inhibitors” OR “Empagliflozin” OR “Dapagliflozin” OR “Canagliflozin”) AND (“Non-alcoholic Fatty Liver Disease”[MeSH] OR “Metabolic Dysfunction-Associated Steatotic Liver Disease” OR “MASLD” OR “NAFLD” OR “Non-alcoholic Steatohepatitis” OR “MASH”) AND (“Carcinoma, Hepatocellular”[MeSH] OR “Hepatocellular Carcinoma” OR “HCC” OR “Liver Cancer”). Titles and abstracts were screened, followed by full-text assessment using predefined inclusion criteria, and data extraction was conducted independently by two reviewers using a standardized template capturing study design, population characteristics, exposure definitions, comparator therapies, and HCC outcomes. The risk of bias of included observational studies was evaluated using the ROBINS-I tool, with domains addressing confounding, selection, intervention classification, missing data, outcome assessment, and reporting bias, and disagreements were resolved by consensus. Quantitative synthesis was performed using R (version 4.4.0) with a random-effects model to account for between-study variability, with pooled risk ratios (RRs) calculated using the inverse variance method and heterogeneity assessed through I², Cochran’s Q, and τ² statistics. Sensitivity analyses were prespecified to explore the robustness of findings by restricting comparisons to studies evaluating SGLT2 inhibitors exclusively against non-SGLT2 oral hypoglycemic agents, allowing a more focused assessment of their potential role in modifying cancer risk within the broader spectrum of metabolic disease.

RESULTS

From an initial pool of 51 records, a structured screening process (Figure 1) yielded three retrospective cohort studies suitable for quantitative synthesis, encompassing a total of 352890 individuals with MASLD, of whom 131282 were exposed to SGLT2 inhibitors and 221608 received alternative oral hypoglycemic therapies. Across these large observational datasets, a consistent signal emerged: SGLT2 inhibitor use was associated with a significantly lower incidence of HCC, with a pooled RR of 0.60 (95%CI: 0.52-0.70; P < 0.0001), corresponding to an approximate 40% relative risk reduction and accompanied by low statistical heterogeneity (I² = 14.1%), supporting internal consistency across populations. This effect persisted in sensitivity analyses restricted to non-SGLT2 comparator therapies (RR = 0.62; 95%CI: 0.45-0.86; P = 0.0038), albeit with moderate heterogeneity (I² = 56.6%), suggesting that the observed association is not solely driven by comparator selection but reflects a broader and reproducible pattern (Figure 2). Although derived from non-randomized data, the convergence of findings across large cohorts, the magnitude of effect, and the stability across analyses collectively point toward a biologically meaningful signal, one that resonates with the emerging paradigm of SGLT2 inhibition as a modifier of downstream complications in metabolic disease and sets the stage for comparison with parallel observations in DKD.

Figure 1 PRISMA flow diagram. HCC: Hepatocellular carcinoma.

Figure 2 Forest plot. A: The association between sodium–glucose cotransporter-2 inhibitor use and hepatocellular carcinoma incidence in adults with metabolic dysfunction–associated steatotic liver disease (primary meta-analysis of three cohort studies); B: The sensitivity analysis restricted to studies comparing sodium–glucose cotransporter-2 inhibitors with non–sodium–glucose cotransporter-2 oral hypoglycemic agents. RR: Relative risk; SGLT2: Sodium-glucose cotransporter-2.

DISCUSSION

The present findings should be interpreted within the broader context of metabolic disease as a multisystem process, in which shared pathways of insulin resistance, inflammation, oxidative stress, and fibrosis drive parallel organ injury. In MASLD, now affecting nearly one-third of the adult population globally[3], these mechanisms underpin progression from steatosis to cirrhosis and HCC, with an increasing proportion of cases arising in both cirrhotic and non-cirrhotic settings across diverse populations[2,3]. Against this background, our analysis-encompassing more than 350000 individuals-demonstrates a consistent association between SGLT2 inhibitor use and reduced HCC incidence, aligning with prior observational and mechanistic data suggesting improvements in steatosis, oxidative stress, fibrosis, and hepatocarcinogenesis[17-26]. Importantly, these biological effects closely mirror the pathophysiological processes implicated in DKD, where persistent proteinuria reflects ongoing inflammatory and fibrotic activity despite optimized therapy. In this regard, the study by Pasari et al[1] provides a complementary and clinically relevant perspective, demonstrating that even with maximally tolerated dapagliflozin and telmisartan, residual renal risk persists and may be attenuated through the addition of finerenone, reinforcing the concept of layered therapeutic strategies targeting complementary pathways within a shared metabolic substrate. Taken together, these observations suggest that SGLT2 inhibitors may act upstream in the metabolic cascade, modifying the underlying disease milieu upon which downstream organ-specific complications-such as renal fibrosis and hepatic carcinogenesis-develop (Table 1).

Table 1 Summary of selected studies.

Ref.	Study type	Follow-up period	Population	Intervention/comparison	Confounders	Conclusions
Cho et al[16], 2024 (Republic of Korea)	Retrospective cohort	Median follow-up 3.56 years	Type 2 diabetes patients with MASLD	SGLT2 inhibitors compared to OGLD	Comorbid viral hepatitis	SGLT2i may be a beneficial option for diabetes management in patients with T2DM, MASLD, and chronic viral hepatitis (risk ratio: 0.54, 95%CI: 0.41-0.70)
Shen et al[17], 2024 (United States)	Retrospective cohort	January 2014-December 2022 (9 years)	Patients with T2DM and MASLD	Metformin and SGLT2 inhibitors compared to DPP4 inhibitors	Concurrent metformin use	No significant difference in the incidence of HCC between the two groups (risk ratio: 0.75, 95%CI: 0.53-1.04)
Mao et al[18], 2024 (United States)	Retrospective cohort	April 2013-December 2021 (8 years)	Patients with T2DM and MASLD	SGLT2 inhibitors compared to other oral hypoglycemic drugs	Other patients were also taking metformin	SGLT2 inhibitor users were associated with a substantially lower risk of HCC than OGLD users among patients with diabetic MASLD (risk ratio: 0.60, 95%CI: 0.49-0.73). The risk was further reduced with concomitant metformin use

SGLT2: Sodium-glucose cotransporter-2; MASLD: Metabolic dysfunction–associated steatotic liver disease; HCC: Hepatocellular carcinoma; T2DM: Type 2 diabetes mellitus; OGLD: Other glucose-lowering drugs.

This paradigm is further supported by a growing body of evidence integrating metabolic, hepatic, and systemic outcomes. Established interventions, including lifestyle modification, metformin, and glucagon-like peptide-1 receptor agonists, have demonstrated benefits in reducing steatosis, fibrosis, and, in selected cohorts, HCC risk, underscoring the multifactorial nature of disease modification in MASLD[19-22,27]. Prospective and population-based studies further indicate that SGLT2 inhibitors improve hepatic steatosis, cardiometabolic profiles, and fibrosis beyond weight loss alone, supporting a metabolically targeted approach to liver disease[28,29]. At the same time, evolving HCC epidemiology and advances in staging and surveillance frameworks-including the Barcelona Clínic Liver Cancer system-highlight the increasing burden of MASLD-related HCC across both cirrhotic and non-cirrhotic populations and the need for improved early detection strategies, including the potential integration of artificial intelligence in imaging pathways[30-35]. Complementary data on sarcopenia, exercise, and bariatric surgery further emphasize that meaningful HCC prevention will likely require multidimensional strategies combining pharmacologic and non-pharmacologic interventions[34-36]. Within this landscape, real-world cohort studies have demonstrated that SGLT2 inhibitor use is associated with reduced hepatic decompensation, mortality, and major liver outcomes even in advanced disease, suggesting clinically meaningful benefits beyond surrogate markers[37], while large network meta-analyses indicate that SGLT2 inhibitors may rank among the most effective antidiabetic classes for reducing HCC incidence, albeit based on indirect comparisons[38]. However, not all analyses are consistent, with some meta-analyses reporting neutral effects on HCC when examined in isolation, highlighting the influence of heterogeneity, residual confounding, and differences in study design[39]. In parallel, cardiovascular and renal trial data demonstrate that combination metabolic strategies-including SGLT2 inhibitors with finerenone or glucagon-like peptide-1 receptor agonists-are feasible, safe, and potentially synergistic, reinforcing the concept of integrated, cross-organ disease modification[40]. Importantly, emerging real-world data in cirrhosis further extend this paradigm, demonstrating that SGLT2 inhibitor use is associated with reduced portal hypertensive complications-including ascites, hepatic encephalopathy, and hepatorenal syndrome-as well as improved survival across etiologies, supporting a clinically meaningful impact even in advanced liver disease[41].

Within this framework, the findings of Pasari et al[1] and the present analysis should be viewed as converging evidence of a broader therapeutic shift: SGLT2 inhibitors are emerging not merely as glucose-lowering agents, but as central components of a disease-modifying strategy capable of attenuating multisystem complications of metabolic syndrome. Nonetheless, important limitations must be acknowledged. The evidence remains derived from observational cohorts, and therefore subject to residual confounding, including the influence of concomitant therapies such as metformin, which may partially explain the increased heterogeneity observed in sensitivity analyses. In addition, variability in MASLD definitions and comparator groups across studies introduces further clinical heterogeneity. The potential for healthy user bias should also be considered, as individuals prescribed SGLT2 inhibitors may differ systematically from controls in terms of healthcare engagement, comorbidity management, and baseline risk. Furthermore, the duration and stage of MASLD prior to SGLT2 inhibitor initiation were inconsistently reported and difficult to control for, which may influence the observed association with HCC. Finally, the long latency of hepatocarcinogenesis limits the ability of currently available studies to fully capture causal effects. These factors collectively necessitate cautious interpretation and underscore the need for well-designed prospective studies with standardized definitions and longitudinal follow-up to clarify the role of SGLT2 inhibitors in modifying liver-related outcomes.

CONCLUSION

In summary, the present analysis, together with the study by Pasari et al[1] supports a converging and increasingly compelling paradigm in which SGLT2 inhibitors extend beyond glucose lowering to function as systemic modifiers of metabolic disease across organ systems. While our meta-analysis suggests a consistent signal toward reduced HCC incidence in individuals with MASLD, this association remains derived from observational data and should be interpreted cautiously; however, when viewed alongside robust evidence of renal and cardiovascular protection-and particularly the demonstration of residual risk modulation through layered therapy with agents such as finerenone in DKD-a coherent biological narrative emerges. Rather than acting as isolated organ-specific therapies, SGLT2 inhibitors appear to target upstream metabolic, inflammatory, and fibrotic pathways that drive parallel disease progression in the liver and kidney, suggesting that their clinical impact may be best understood as attenuation of the substrate from which downstream complications, including fibrosis and carcinogenesis, arise. In this context, the observed reduction in HCC risk should not be interpreted as definitive chemoprevention, but as part of a broader, multidimensional effect on metabolic disease biology, likely modest, indirect, and synergistic with other interventions. This integrated perspective reframes both datasets-not as isolated findings, but as complementary signals of a therapeutic shift toward cross-organ disease modification. Accordingly, future research should move beyond traditional single-organ endpoints and prioritize prospective, multidomain trials designed to evaluate the impact of combined metabolic therapies on renal, cardiovascular, and hepatic outcomes simultaneously, thereby determining whether modulation of the metabolic milieu can meaningfully alter the natural history of complex diseases such as MASLD (e.g., liver stiffness using fibroscan) and DKD.

ACKNOWLEDGEMENTS

We extend our appreciation to the Faculty of Life Sciences and Education at the University of South Wales in association with Learna Ltd. for the Gastroenterology MSc program and their invaluable support in our work. We sincerely acknowledge the efforts of the University of South Wales and commend them for their commitment to providing life-long learning opportunities and advanced life skills to Healthcare professionals.