Introduction
According to global cancer statistics for 2022, ~20 million new cancer cases were diagnosed worldwide, resulting in 9.7 million deaths. Projections suggest that the number of cancer cases will rise to 35 million by 2050, reflecting population growth and aging trends (1). These estimates highlight the serious threat posed by malignant tumors to global health, as well as their substantial social and economic impact. Therefore, improving cancer treatment is not only a medical priority but also a crucial public health objective. Cancer immunotherapy has emerged as a transformative approach that mobilizes the patient's own immune system to recognize and eliminate malignant cells (2,3). Major immunotherapeutic strategies, including immune checkpoint inhibitors (ICIs), adoptive cell transfer and cancer vaccines, have made substantial advancements in the treatment of various diseases and have provided patients with advanced cancer with the possibility of a long-term life (4,5). However, the broader application of these therapies is limited by issues such as primary or acquired resistance, limited response rates and immune-related adverse events. There is a pressing need to identify novel strategies that can enhance the efficacy and expand the applicability of cancer immunotherapy (6,7).
Metformin, a first-line biguanide agent for type 2 diabetes mellitus, is widely recognized for its favorable safety and tolerability profile. It improves glycemic control through multiple mechanisms: Reducing intestinal glucose absorption, decreasing hepatic glucose output (gluconeogenesis) and enhancing insulin sensitivity in peripheral tissues (8–10). Beyond its glucoregulatory actions, metformin has garnered significant interest in oncology over the past decade. A growing body of preclinical evidence suggests that metformin may possess direct and indirect antitumor properties and could potentially augment the efficacy of cancer immunotherapy (11–14). It may exert an anti-tumor effect via activating the adenosine monophosphate-activated protein kinase (AMPK)/mammalian target of rapamycin (mTOR) pathway (12), decreasing programmed death ligand 1 (PD-L1) expression (15), suppressing mitochondrial complex I activity (16) and modulating the tumor microenvironment (TME) (17,18). Numerous clinical studies have shown its association with a lower risk and better prognosis for cancer patients (19–21). It has also been shown to work in concert with immunotherapy (22,23), establishing a theoretical basis for the concurrent use of metformin and immunotherapy (24–26).
Despite these promising preclinical and pharmacological insights, the clinical efficacy of combining metformin with immunotherapy remains controversial (27,28). Therefore, a meta-analysis was performed in the present study to synthesize the available evidence regarding the association between metformin use and survival outcomes in cancer patients receiving ICIs, with the aim of clarifying its potential role as an adjunctive therapy in immunotherapy.
Discussion
The present meta-analysis, encompassing 9 studies and 5,014 cancer patients, investigated the association between concomitant metformin use and survival outcomes in patients receiving ICIs. The pooled results demonstrate that metformin use is significantly associated with improved PFS (HR=0.80, 95% CI: 0.69–0.93, P=0.005) and OS (HR=0.78, 95% CI: 0.62–0.97; P=0.027) in this patient population. However, subgroup analyses revealed a striking geographical disparity: The survival benefit was statistically significant and pronounced in Asian populations (OS: HR=0.66, 95% CI: 0.47–0.91, P=0.012; PFS: HR=0.69, 95% CI: 0.58–0.82, P<0.001) but was not observed in non-Asian cohorts. These findings suggest a complex interaction between metformin and immunotherapy, potentially modulated by regional factors.
Previous meta-analyses, including the study by Wen et al (40), evaluated metformin across various anticancer treatment modalities. In contrast, the present analysis focused specifically on ICI-based therapies, thereby offering a more immunotherapy-oriented perspective. In addition, compared with the meta-analysis conducted by Shen et al (41), which reported no significant survival benefit and suggested a potential unfavorable effect of metformin on overall survival, the present study incorporated additional clinical studies published up to 2024 and applied stricter inclusion criteria by excluding conference abstracts, letters, studies lacking complete survival data and those with lower methodological quality, which may enhance data completeness and analytical reliability. By contrast, the pooled analysis performed in the present study demonstrated a statistically significant improvement in both OS and PFS with metformin use in combination with ICIs. The present subgroup analysis further provides meta-analytic evidence suggesting a potentially greater survival benefit among Asian patients. This regional trend was not clearly delineated in prior studies and may offer new insights into population-based stratification strategies in precision immunotherapy. Furthermore, this observation raises the possibility that differences in metabolic phenotype, genetic background or treatment patterns may interact with metformin-mediated immunomodulation.
The observed OS benefit aligns with the premise derived from preclinical studies that metformin can potentiate anti-tumor immunity (42–44). Metformin can inhibit the proliferation and metabolism of tumor cells by activating the AMPK signaling pathway and inhibiting mTOR signaling (45). It can also reduce the expression of PD-L1 and block PD-1/PD-L1 signaling, thereby enhancing the anti-tumor activity of T cells (46). Metformin may also enhance immune-cell activity and regulate the TME. A theoretical foundation for its combination with immunotherapy is provided by the aforementioned findings. Turpin et al (16) published an article in 2024 examining the utilization of patient-derived explant culture (PDEC) to investigate the tumor immune microenvironment (TIME), employing this model to assess the impact of anti-tumor agents, including the combination of venetoclax and metformin, on immune-cell functionality. The study's findings indicated that the PDEC model demonstrated that metformin activated dendritic cells within the TIME by inhibiting mitochondrial respiratory chain complex, thereby augmenting the anti-tumor immune response of CD4+ T cells, and underscored the significance of the PDEC model in investigating the TIME and the mechanisms of drug action. Tan et al (47) assessed the anti-tumor efficacy of combining PD-1 inhibitors with mTOR inhibitors rapamycin or metformin in triple-negative breast cancer (TNBC). According to their data, metformin and rapamycin may both significantly lower PD-L1 expression and inhibit mTOR pathway activity in TNBC. The combination of PD-1 inhibitors with these agents markedly decreased tumor growth and metastasis, increased CD8+ T-cell infiltration and tumor-cell apoptosis, and amplified the anti-tumor efficacy of PD-1 inhibitors. Wabitsch et al (26) revealed that non-alcoholic steatohepatitis (NASH) reduces the efficacy of ICI therapy for liver cancer by impairing the metabolism and migration capabilities of CD8+ T cells. In NASH mice, metformin therapy may enhance CD8+ T-cell metabolic activity, regaining the effectiveness of ICI treatment. These fundamental study results provide a strong theoretical foundation for metformin's possible use in cancer immunotherapy. Building on this mechanistic understanding, a recent study on TNBC has demonstrated that low-dose metformin activates the AMPK-acetyl-CoA carboxylase-fatty acid β-oxidation signaling axis, thereby inducing Src kinase activation and enhancing anti-tumor immunity, whereas high-dose metformin suppresses this pathway and may even exert tumor-promoting effects (48). However, critical details regarding metformin dosage, treatment duration, timing of metformin initiation relative to ICI therapy (before, concurrent with or after ICI), and steady-state blood concentration in patients were unavailable in the included studies, limiting further exploration of optimal therapeutic parameters. Future prospective studies designing combination therapies should emphasize dose optimization and clearly report the timing of metformin initiation relative to ICI to maximize efficacy while minimizing toxicity.
Additionally, studies indicate that diabetes can alter the immune landscape within the TME of solid tumors, potentially fostering an immunosuppressive state (49). The immunomodulatory benefits of metformin, particularly its potential to enhance antitumor immunity, may be closely linked to the improvement of diabetic metabolic conditions. This interplay warrants further investigation through well-controlled animal studies employing both diabetic and non-diabetic models to dissect the specific contributions of metabolic normalization vs. direct drug effects. Beyond metabolic factors, substantial evidence suggests that heterogeneity in the TME may lead to differential efficacy. For instance, brain tumors and uveal melanomas, characterized by minimal tumor-infiltrating immune cells and a macrophage-dominated milieu, often exhibit resistance to ICIs (50–52). In contrast, malignancies such as lung adenocarcinoma, head and neck squamous cell carcinoma and cutaneous melanoma, which typically display richer immune cell infiltration, tend to respond more favorably to immunotherapy (53–55). Future multi-center studies with large-sample cohorts for individual cancer types are warranted to explore the impact of metformin on immunotherapy outcomes across different tumor sites.
The most intriguing finding of the present analysis is the significant disparity in treatment effect between Asian and non-Asian populations. The reasons for this heterogeneity are likely multifactorial and may involve differences in tumor biology, genetic predisposition, pharmacogenomics and clinical practice patterns. The lack of a statistically significant benefit in non-Asian cohorts underscores the potential influence of divergent genetic backgrounds, lifestyles or clinical management approaches. Future research should integrate multi-omics data, including genomics and metabolomics, to elucidate the underlying mechanisms, and conduct multicenter trials to validate and refine population-specific therapeutic strategies. The study by Bouchi et al (56) highlighted that the pathophysiology of diabetes, levels of obesity and insulin resistance among patients vary across Eastern and Western populations, along with disparities in pharmacological choices and treatment methodologies. Lin et al (57) utilized genome-wide association analysis, cross-ancestry meta—analysis and Mendelian randomization analysis to explore the genetic structure of metabolites in Han Chinese and European populations and their correlation with diseases. They revealed the genetic differences in metabolites between these ethnic groups and their impact on complex diseases, underscoring the importance of ethnic diversity in genetic research. These recent advances in metabolic research also provide unique insights into the present subgroup analysis results. The metabolic genetics, pharmacological sensitivities, immunological responses and other characteristics of individuals from different locations may be very different from one another. The efficiency of immunotherapy in combination with metformin metabolic processes may be impacted by changes in dietary and lifestyle practices. Subsequent research should account for additional personalized variables and investigate its mechanism of action in more depth.
The present meta-analysis has several limitations. Selection bias may exist, since all of the included studies are retrospective cohort studies. The lack of individual patient data prevented more detailed analyses, such as examining the impact of gene expression or metabolic state on the outcomes. In conclusion, it may be suggested that metformin may enhance the anti-tumor efficacy of immunotherapy by collaborating with it, offering a theoretical foundation for future combination treatment strategies that emphasize immune modulation.
In conclusion, this meta-analysis provides evidence that metformin use is associated with enhanced survival outcomes in cancer patients treated with immune checkpoint inhibitors, with a particularly significant effect observed in Asian populations. These results underscore the potential of metformin as an inexpensive, widely available and generally well-tolerated agent to augment cancer immunotherapy. However, the geographical heterogeneity in treatment response cautions against its broad, indiscriminate use and emphasizes the necessity for predictive biomarkers. Future efforts should be directed toward prospective, randomized controlled trials specifically designed to validate the efficacy of metformin and ICI combination therapy. These trials should prioritize biomarker-driven patient selection, potentially focusing on individuals whose tumors exhibit specific metabolic vulnerabilities or who belong to ethnic subgroups most likely to benefit.