Action Mechanism of Metformin and Its Application in Hematological Malignancy Treatments: A Review

Hematologic malignancies (HMs) mainly include acute and chronic leukemia, lymphoma, myeloma and other heterogeneous tumors that seriously threaten human life and health. The common effective treatments are radiotherapy, chemotherapy and hematopoietic stem cell transplantation (HSCT), which have limited options and are prone to tumor recurrence and (or) drug resistance. Metformin is the first-line drug for the treatment of type 2 diabetes (T2DM). Recently, studies identified the potential anti-cancer ability of metformin in both T2DM patients and patients that are non-diabetic. The latest epidemiological and preclinical studies suggested a potential benefit of metformin in the prevention and treatment of patients with HM. The mechanism may involve the activation of the adenosine monophosphate-activated protein kinase (AMPK) signaling pathway by metformin as well as other AMPK-independent pathways to exert anti-cancer properties. In addition, combining current conventional anti-cancer drugs with metformin may improve the efficacy and reduce adverse drug reactions. Therefore, metformin can also be used as an adjuvant therapeutic agent for HM. This paper highlights the anti-hyperglycemic effects and potential anti-cancer effects of metformin, and also compiles the in vitro and clinical trials of metformin as an anti-cancer and chemosensitizing agent for the treatment of HM. The need for future research on the use of metformin in the treatment of HM is indicated.


Introduction
Hematological malignancies (HMs) mainly include heterogeneous tumors, such as leukemia, lymphoma and myeloma [1,2], and the mechanisms are abnormal hematological cells or immune cells with blocked differentiation and indefinite proliferation followed by biological organismal dysfunction [3]. HMs are different from solid tumors, and due to their specificity, conventional treatment options mainly include chemotherapy, radiotherapy and hematopoietic stem cell transplantation (HSCT) [4]. Although the current conventional first-line treatment has a certain efficacy, primary or secondary drug resistance still often occurs, resulting in a poor overall efficacy with respect to the tumor [5]. Therefore, the search for new therapeutic targets and the development of new treatments remains an important operation with scientific significance in the research of HMs [6].
Metformin is recognized by the World Health Organization (WHO) as an essential drug for the treatment of patients with diabetes and can be widely used in the treatment of type 2 diabetes (T2DM) alone or in combination with other drugs [7]. The most classic action of metformin is to lower hyperglycemia and reduce its clinical symptoms [8]. Due to its low cost and high clinical value in terms of glycemic control and safety, it is now one of the most prescribed drugs in the market [9].
In 1995, Evans et al. first discovered that metformin not only controlled blood glucose in patients with diabetes, but also reduced the risk of malignancies [ [33,34]. Therefore, this review focuses on the mechanism of action of metformin in cancer prevention and the potential role of this drug in the treatment of patients with HM.
In addition, several studies have supported the therapeutic benefit of metformin in patients with HM. Preclinical data, clinical trial reports, reviews and meta-analyses have reported the effect of metformin as an anti-cancer agent in HM from various perspectives, but these articles reached different conclusions. In this paper, we further focus on the specific mechanisms involved in metformin in cancer treatments, starting with the antihyperglycemic effect action mechanism of metformin. Focusing on a collection of published studies on the anti-cancer mechanism of the action of metformin in HM and clinical studies, we affirm the therapeutic potential of metformin for future use in patients with HM.

Metformin Mechanism of Action
Metformin, a natural derivative of guanidine present in French cloves [35], was first reported to be synthesized by Werner and Bell in 1922 [36,37]. Metformin was first introduced as an anti-hypoglycemic effect agent in the treatment of diabetes by Jean Sterne in 1957 [35], and the following decades saw intense medical interest in the effects of metformin. In 1994, the U.S. Food and Drug Administration (FDA) approved the use of metformin for the treatment of T2DM [38]. In 1998, the UK Prospective Diabetes Study (UKPDS) confirmed the safety and efficacy of metformin as an initial treatment for hyperglycemia in patients with T2DM [39]. This opened an era of using metformin as the first-line treatment for T2DM.
Recent studies have shown that the gastrointestinal tract and its intestinal microorganisms are also involved in the anti-hyperglycemic effects of metformin [53,54]. Metformin increases the intestinal absorption and utilization of glucose in the intestinal cells, which in turn limits the entry of glucose into the bloodstream [55,56]. Altering the composition of the intestinal flora and the secretion of certain key molecules, such as glucagon-like peptide-1 (GLP1) and growth differentiation factor 15 (GDF15), during treatments can affect the gastrointestinal microenvironment and produce beneficial metabolic effects [57,58]. Figure 1 summarizes the mechanisms described above. Zhou et al. demonstrated that metformin can activate adenosine monophosphateactivated protein kinase (AMPK) in isolated rat hepatocytes, providing a cellular mechanism for the anti-hyperglycemic effect of metformin [59]. AMPK is an energy sensor that acts in numerous metabolic pathways involved in intracellular energy homeostasis [60].
The activation of AMPK involves many different pathways. In 2000, two independ- It is commonly believed that metformin exerts its anti-cancer activity via the LKB1-AMPK-mTOR pathway, and the overexpression of mTOR in this pathway is often associated with the development of numerous diseases such as tumors, among others [86][87][88][89][90]. Metformin activates AMPK after the inhibition of ETC I, inhibits the expression of mTOR and further inhibits the expression of more important factors mediating downstream procarcinogenic pathways and the process of tumor development, such as nuclear factor kappa B (NF-κB), interleukin-6 (IL6), mitogen-activated protein kinase (MAPK), Ras and c-MYC [91][92][93]. In animal or in vitro model experiments, high doses (~mM) of metformin inhibited mitochondrial respiratory chain complex I, leading to an increase in the AMP/ATP ratio and prompting the phosphorylation and activation of the upstream kinase LKB1, while low doses of metformin (~µM) activated AMPK via the lysosomal pathway [65]. The mammalian target of rapamycin complex 1 (mTORC1) activation by metformin in the lysosomal pathway via the Ras-associated guanosine triphosphatase (GTPase) completes the regulation of AMPK, which is independent of the AMP/ATP ratio [65,[94][95][96]. The activation of the LKB1-AMPK-mTOR signaling pathway can usually mediate the following cellular responses.
1. Regulation of epigenetic modifying enzyme activity [97]: AMPK directly and indirectly regulates histone acetylation and alters gene expression patterns via the epigenetic regulation of various chromatin functions [98,99]. Activated AMPK phosphorylates a variety of substrates including histone acetyltransferases (HATs), histone deacetylases (HDACs) and deoxyribonucleic acid (DNA) methyltransferases (DNMTs), often leading to their inhibition [98][99][100]; 2. Activation of p53: Activated P53 promotes apoptosis, cycle arrest, autophagy and the inhibition of the phosphoinositide 3-kinase (PI3K)-protein kinase B (AKT) signaling pathway [101,102]. The effect of P53 in tumor apoptosis was also associated with the reduced expression of transcription factor specificity protein 1 (Sp1), Sp3, Sp4 and oncogenes (B-cell lymphoma-2 (BCL-2), mTOR, vascular endothelial growth factor (VEGF) and MYC, etc.) [103,104]; 3. Activation or inhibition of the expression of other cancer-related signaling pathways: 3.1 Activation of the AMPK-forkhead box O1/3 (FOXO1/3) signaling pathway [105][106][107]: FOXO is a proliferation-associated transcription factor, and the activation of FOXO3a has been shown to be required for the pro-apoptotic and chemotherapeutic effects of metformin in a variety of tumor models [108,109]; 3.2 Suppression of Hippo pathway expression [110,111]: In some (cancer) models, metformin has been shown to repress the activity or expression of yes-associated protein (YAP)/tea domain transcription factor (TAZ), which are two central effectors of the Hippo pathway that are mediated by AMPK [112][113][114][115][116][117] [126]. Subsequently, metformin activates additional signaling pathways for targeting CSCs via the activation of AMPK, which include PI3K-AKT-mTOR, insulin-insulin growth factor1 (IGF1), MAPK, sonic hedgehog (Shh), Wnt, TGFB, Notch, Hippo and NFkB pathways [126]; 5. Reducing IRS1 phosphorylation [127]: activated AMPK inhibits hyperinsulinemiaassociated tumor activity by reducing the circulating insulin levels and targeting the insulin-IGF1-PI3K signaling axis [127,128]; 6. Mitigation of hypoxia and other tumor responses caused by hypoxia: Metformin leads to slow tumor growth by decreasing the expression of hypoxia-inducible factor-1α (HIF-1α), which in turn decreases the expression of HIF1-targeted genes [129][130][131][132]. The inhibition of HIF-1α simultaneously suppresses the immunosuppressive activity of myeloid-derived suppressor cells (MDSCs) and improves T-cell immunity in the tumor microenvironment (TME) [133][134][135]. These immune responses are ultimately also involved in the inhibition of angiogenesis [136]; 7. Inhibition of adipogenesis: The cancer cells themselves require more nutrients and energy; thus, the rate of adipogenesis is higher [137]. AMPK inhibits lipogenesis by targeting the activity or expression of many lipogenic enzymes. On the one hand, AMPK phosphorylates and inhibits 3-hydroxy-3-methylglutaryl CoA reductase (HMGCR), which catalyzes the rate-limiting step in cholesterol synthesis. On the other hand, AMPK phosphorylates and inactivates acetyl CoA carboxylase (ACC), the main enzyme involved in the biosynthesis of fatty acid and HMGCR, resulting in the inhibition of cholesterol biosynthesis [138]. In addition, AMPK also phosphorylates sterol regulatory element binding protein-1c (SREBP-1c) at Ser-372, which restricts its cleavage and nuclear translocation. This process results in the downregulation of SREBP-1c target genes, including those encoding ACC1 and fatty acid synthase (FASN), and leads to reduced lipogenesis [139]; 8. Activation of ataxia telangiectasia-mutated gene (ATM), which leads to the activation of DNA damage repair pathways and the inhibition of tumor growth [140].

Activation of AMPK-Independent Signaling Pathways
Some of the pathways controlled by metformin can also be modified in a manner that is not dependent on AMPK activation. Metformin can promote crosstalk between apoptosis and autophagy in multiple ways to accomplish anti-tumor effects [141]. These include downregulating the signal transducer and the activator of transcription 3 (STAT3)/Bcl2/Beclin1 signaling pathway activity in cancer cells, disrupting unfolded protein response (UPR) transcription, blocking the oncogenic Wnt signaling pathway, regulating caspase enzyme activity, increasing reactive oxygen species (ROS) production by cancer cells, inhibiting the activation of mTORC1 by Rag GTPases and inhibiting glutaminase activity [142][143][144][145][146][147][148][149]. Shen et al. found the preferential inhibition of cell-cycle-related proteins among the broad inhibitory effects of metformin on protein synthesis [150]. Specifically, the metformin activation of the p53-regulated in the development DNA damage response 1 (REDD1) axis caused an independent inhibition of AMPK by mTOR, leading to a decrease in cell-cycle protein D1 (cyclin D1) expression, which in turn inhibited the G0/G1 cell cycle [66,151]. In addition, metformin induces the expression of dicer, an enzyme that processes microRNAs (miRNAs) precursors into mature miRNAs, as a means of regulating miRNAs to exert anti-cancer effects [152]. Numerous experiments have shown that metformin inhibits tumorigenesis and tumor growth by altering the expression of specific miRNAs in human breast, pancreatic, prostate, renal cell, lung, hepatocellular, bile duct, small intestine and human oral cancers [153][154][155][156][157][158][159][160][161][162]. The repression of different miRNAs can directly or indirectly induce apoptosis, cycle arrest and other phenomena. Metformin reduces stressor-induced ROS production, activates endogenous repair systems, prevents ROS toxicity, protects cells and ameliorates genomic instability and possible cancer risk [163]. Metformin reduces the expression of the P-glycoprotein (P-gp) encoded by multidrug resistance1 (MDR1), and, via this action, it blocks drug efflux from cancer cells, allowing metformin to be used as an adjuvant relative to classical chemotherapy [164].

Altered Energy Metabolism of Tumors
Cancer cells have a different pathway for energy production. Even under aerobic conditions, the tricarboxylic acid cycle (TCA cycle) and OXPHOS are restricted, and they favor lactic acid fermentation, called the "Warburg effect" [165,166]. Many studies observed that mitochondrial OXPHOS is the main source of ATP in most cancer tissues [167]. Metformin can reduce the availability of glucose to cancer cells by causing a metabolic shift from OXPHOS to glycolysis via the inhibition of mitochondrial function [61]. Metformin targets mitochondrial glycerol 3-phosphate dehydrogenase (mGPDH), causing the inhibition of mitochondrial function while impairing the production of dihydroxyacetone phosphate (DHAP), which can reduce the process of glucose production from glycerol [80]. In addition, metformin targets mitochondrial integrity by modulating calcium flux in cancer cells [168]. Inducing endoplasmic reticulum stress and releasing calcium into the cytoplasm of cells induces higher calcium uptake by mitochondria, ultimately leading to mitochondrial swelling.

Improved Immune and Inflammatory Response
Inflammatory cells are essential players in the tumor process, promoting the proliferation, survival and migration of tumor cells [169]. In addition to targeting cancer cells, metformin has been shown to target immune cells in the tumor microenvironment, such as CD8 + T cell, regulatory cells (Tregs), MDSCs and tumor-associated macrophages (TAMs), which may contribute to the expression of anti-tumor activities [170].
The NF-κB signaling pathway is a key regulator of immunity and plays an important role in tumor progression. Metformin inhibits the NF-κB signaling pathway and reduces the levels of inflammatory cytokines including tumor necrosis factor-α (TNF-α) and IL-6 [171]. In addition, metformin inhibits ETC I, further inhibiting ROS and interleukin-1β (IL-1β) production via lipopolysaccharide (LPS)-activated macrophages [172]. Metformin can directly and indirectly alter the function of CD8 + T cells, blocking T cell depletion by increasing the number of CD8 + T lymphocytes (CD8 + TILs) and protecting them from apoptosis and exhaustion [173,174]. In addition, metformin can reduce programmed cell death protein 1 ligand 1 (PD-L1) in tumor cells via multiple pathways, leading to enhanced cytotoxic T lymphocyte (CTL) activity [175]. In addition, metformin helps the body produce sustained anti-tumor immunity by acting on multiple immune pathways.

Therapeutic Application of Metformin in Hematological Malignancies
A large number of different types of studies have shown that metformin is a potential therapeutic target for patients with leukemia, lymphoma and multiple myeloma, and it is expected that the pleiotropic effects of this drug could act on multiple targets [53, [187][188][189][190][191][192][193][194]. Herein, we discuss experimental ex vivo studies and clinical observations of metformin in different hematologic malignancies. We compiled tables of preclinical applications of metformin in malignant hematological during the last decade, divided into in vivo animal experiments (Table 1), in vitro cellular experiments in leukemia (Table 2), lymphoma (Table 3) and myeloma (Table 4).

Therapeutic Application of Metformin in Hematological Malignancies
A large number of different types of studies have shown that metformin is a potential therapeutic target for patients with leukemia, lymphoma and multiple myeloma, and it is expected that the pleiotropic effects of this drug could act on multiple targets [53, [187][188][189][190][191][192][193][194]. Herein, we discuss experimental ex vivo studies and clinical observations of metformin in different hematologic malignancies. We compiled tables of preclinical applications of metformin in malignant hematological during the last decade, divided into in vivo animal experiments (Table 1), in vitro cellular experiments in leukemia (Table 2), lymphoma (Table 3) and myeloma (Table 4). Metformin induced AMPK activation, mTOR inhibition and remarkably blocked tumor growth in murine lymphoma xenografts. [193] i   The combination of 6-BT with metformin resulted in significant cytotoxicity (60-70%) in monocytic AML cell lines and was associated with the inhibition of FLT3-ITD activated STAT5 and reduced c-Myc and GLUT-1 expression. [218] OCI-AML3, REH, NALM-6-6, and KBM5 cells AML ALL CML 0-10 mM, 0-16 h 2016 ABT-737 Inhibition of mitochondrial metabolism by metformin or phenformin was associated with increased leukemia cell susceptibility to induction of intrinsic apoptosis.  In the presence of metformin, the anti-cancer potential of sorafenib, accompanied by increased LC3 levels, was found to be synergistically enhanced with the remarkably reduced protein expression of the mTOR/p70S6K/4EBP1 pathway, while not appreciably altering the cell cycle.
[221] Metformin mediated anti-leukemia effects by proapoptosis and inhibition of mTORC1 signaling, potentiating the anti-cancer efficacy of imatinib in Ph+ ALL cells. Metformin-induced autophagy was associated with the activation of the ERK pathway.     Metformin-induced AMPK activation was associated with the inhibition of mTOR signaling without involving AKT. [193] 2DG  Metformin inhibited GRP78 to enhance the anti-myeloma effect of bortezomib. [194]  Ritonavir and metformin effectively suppressed AKT and mTORC1 phosphorylation and prosurvival BCL-2 family member MCL-1 expression in multiple myeloma cell lines. AML is the most common acute leukemia in adults and is characterized by the uncontrolled proliferation of leukemic cells that inhibit normal bone marrow hematopoiesis following varying degrees of differentiation blockage with respect to bone marrow progenitor cells [239].
Since metformin was shown to interfere with the proliferation and clonal activity of AML cells without affecting normal CD34 + HSCs, numerous experiments have begun to further explore the anti-AML effects of metformin [240]. Metformin accelerates AML cell apoptosis by inhibiting OXPHOS, decreasing FOXM1, activating the LKB1-AMPK signaling pathway and dephosphorylating eukaryotic initiation factor 4E binding protein 1 (4E-BP1) [213,224,241,242]. In acute progranulocytic leukemia (APL) cell lines, metformin induced c-Myc degradation and apoptosis via the MEK-ERK signaling pathway [243]. In addition, metformin combined with sorafenib, cytarabine and venetoclax showed synergistic enhancement effects in AML treatment, and these effects were demonstrated by mTOR inhibition and the regulation of apoptosis-related proteins [196,199,221]. Clinically, a retrospective hospital cohort study found that although metformin users did not perform better than non-users in overall survival (OS) and disease-free status, they performed better than insulin and oral diabetes medication-treated AML patients [244].

Chronic Myeloid Leukemia (CML)
CML is characterized by the presence of the Philadelphia (Ph) chromosome with a reciprocal translocation between chromosomes 9 and 22, and this translocation generates the BCR/ABL fusion gene [245]. A tyrosine kinase is produced by the BCR/ABL fusion gene, which drives tumorigenesis and prevents apoptosis by activating downstream signals [246].
After AMPK activation, metformin inhibited the proliferation and clonal activity of different CML cell lines, including the T315I BCR-ABL mutant CML line expressing imatinib resistance [247]. With regard to drug combinations, experiments have shown that the combination of nilotinib and metformin is more effective than the combination of nilotinib and c-Jun N-terminal kinase (JNK) inhibitors [208].
An observational study of patients with newly diagnosed chronic-phase CML showed that the proportion of patients reaching complete cytogenetic response (CCyR) was 100% and 73.6% in the metformin and non-metformin groups, respectively, with similar numbers reaching major molecular response (MMR) and complete molecular response (CMR) between the two groups. The median time for achieving MMR and CMR was shorter in the metformin group compared to the non-metformin group (11.1 months vs. 19.5 months; 37.4 months vs. NR), indicating that metformin used together with TKI therapy may increase the proportion of CML patients achieving CCyR and shorten the time to MMR and CMR [248].

Acute Lymphocytic Leukemia (ALL)
ALL is aggressive, with the clonal proliferation of immature lymphocytes at different stages of differentiation that, if not controlled, can cause a destructive infiltration of the bone marrow [249]. In adults, 75% of cases develop from the precursors of the B-cell lineage and the rest consist of malignant T-cell precursors [250].
Metformin increases the chemosensitivity of ALL by inhibiting the AKT-mTOR pathway via the activation of AMPK [251,252]. In T-ALL cells, metformin-stimulated AMPK inhibited mTOR to trigger cellular autophagic responses by a mechanism involving miRNA-19 overexpression [225,253]. Metformin-triggered apoptotic effects also involve the AMPKdependent activation of the endoplasmic reticulum (ER) stress/UPR cell death pathway and the apoptotic mediators including inositol-requiring enzyme1 (IRE1) and C/EBPhomologous protein (CHOP) [225]. In addition to AMPK, the expression of other important metabolic kinases is also regulated by metformin in ALL cells. Metformin regulates the expression of protein kinase C-ε(PKCε) and δ in responding cells, leading to the downregulation of PKCδ and control of PKCε [254]. The altered balance between PKCε and PKCδ has important implications for tumor cell survival, as both kinases control the nuclear factor erythroid 2-related factor (Nrf2) transcription factor [255]. In adult T-cell leukemia (ATL), metformin activates LKB1-AMPK and inhibits leukemic proliferation by reducing Tax expression [256]. Leukemic stem cells (LSCs) are rare cells with leukemic origins that have an intrinsic resistance mechanism relative to chemotherapy [257]. In T-ALL cells, metformin can target the Hoescht 33342 low side population and the CD34 + CD7 − CD4 − subpopulation, which are rich in LSCs [251].
In ALL, metformin can be used in combination with anthracyclines (doxorubicin and adriamycin), dexamethasone, imatinib, proviral integration sites for Moloney murine leukemia virus (PIM)-1/2 kinase inhibitors, thiazolidinediones and other drugs to exert synergistic effects, reducing tumor cell growth and survival by stimulating various signaling pathways and showing a cumulative effect [193,203,222,225,251,252]. The use of metformin in ALL treatment can help reduce the dose of adriamycin that is necessary to prolong remission and reduce the cardiotoxicity of anthracyclines [258]. There are, of course, different phenomena. Studies have shown that the combination of metformin and cisplatin exerts an antagonistic effect in ALL, suggesting that metformin and oncology drugs are not synergistic in one way or the other [206].
In a prospective study of 102 patients with new-onset Ph − B-cell leukemia (NCT03118128), metformin combined with chemotherapy was shown to be effective in ALL patients with elevated ABCB1 gene expression levels (p = 0.025), and in the metformin user group the drug was protective against treatment failure (odds ratio (OR) 0.07, 95% confidence interval (CI) 0.0037-1.53) and early relapse (OR 0.05, 95% CI 0.0028-1.153) in the metformin user group [259]. A phase 1 study (2018) exploring metformin in combination with induction chemotherapy for relapsed/refractory (R/R) ALL showed that the combination of metformin and vincristine, dexamethasone, PEG-asparaginase and doxorubicin (VXLD) was tolerable in the 14 included patients; the overall treatment induced ER stress; and toxic reactions occurred at all dose levels of metformin [260]. In the randomized controlled trial of an intensive insulin regimen in patients with hyperglycemic ALL (NCT00500240), the trial was terminated early, although this ALL clinical outcome was invalid. However, a secondary analysis showed that the use of metformin and/or thiazolidinedione was associated with increased progression-free survival (PFS) compared to other drug combinations and may improve the patient prognosis [261]. In a small phase I clinical trial in patients with ALL, the addition of metformin to standard chemotherapy was well tolerated. The unselected use of metformin during this treatment was the only variable considered to have a significant adverse prognosis (p = 0.55). The Cox regression showed that the addition of metformin reduced the risk of recurrence by 56% [262]. A retrospective analysis reported on the safety of the metformin treatment for hyperglycemia control in 17 patients with ALL during the induction period. The data showed no significant toxicity of metformin in all patients, with only one patient failing to control glycemia, suggesting that metformin can be used in clinical ALL patients for glycemic control [263].

Chronic Lymphocytic Leukemia (CLL)
CLL is a common leukemia in western countries, and metformin has been found to be cytotoxic to both resting and activated CLL cells [264]. The cytostatic effect of metformin on activated CLL cells was manifested by the reduced expression of proliferation and proliferation-related molecules. In these cells, metformin inhibited the activation-induced upregulation of chemokine receptors and adhesion molecules that can synergistically stimulate mitogenic activity and cell homing in lymphoid tissues [265]. Metformin inhibits the activation of transcription factors in the CLL pro-survival and pro-activation pathways by activating AMPK and reducing glucose metabolism, further impeding CLL cell survival, activation and clonal expansion [265][266][267][268][269][270]. CLL cells are more dependent on the protective activity of BCL-2 anti-apoptotic protein than normal cells. This also explains the higher sensitivity of CLL cells to metformin than normal lymphocytes [271]. Metformin modulates the glycolytic capacity of CLL cells and limits the availability of extracellular glucose, thereby increasing the sensitivity of CLL cells relative to the cytostatic effects of metformin [264]. The combination of classical anti-CLL drugs fludarabine, ritonavir and ABT-737 with metformin may enhance its cytotoxicity by altering mitochondrial activities and other mechanisms of action, and can exert synergistic effects [53,272].

Multiple Myeloma (MM)
MM is a malignant plasma cell proliferative disease with major symptoms including elevated blood calcium levels, renal insufficiency, anemia and bone damage (CRAB) [273]. In vitro and in vivo experiments demonstrated a direct anti-tumor effect of metformin on myeloma [92,191,194,201,202,235,237,274].
Metformin inhibits the proliferation of myeloma cells by stimulating apoptosis and cellcycle arrest by a mechanism involving the dual inhibition of the AMPK activation-mediated mTORC1 and mTORC2 pathways [201]. In terms of cell proliferation, metformin inhibits the expression of a range of signaling pathways such as IGF1, PI3K, AKT and downstream mammalian rapamycin [190]. Metformin reduces insulin and IGF1 production, increases insulin sensitivity and is thought to reduce the progression of monoclonal gammopathy of undetermined significance (MGUS) relative to multiple myeloma [275,276]. Metformin specifically reduced IL-6R expression via AMPK, is mediated by mTOR and showed synergistic effects with common anti-myeloma drugs [92]. Hypoxia stimulates the migration and homing of malignant plasma cells relative to new bone marrow niches, promoting osteolytic bone destruction and MM-induced angiogenesis [277][278][279][280]. Metformin prevents HIF-1α accumulation in hypoxia and slows disease progression by inhibiting the activity of ETC I [131,281,282]. In addition, cellular experiments demonstrated that metformin could exert anti-MM cellular effects by downregulating the STAT3 signaling pathway; activating the ERK1/2, MAPK and NF-κB pathways; blocking glucose-regulated protein 78 (GRP78) promoter reporter activation; inhibiting UPR activation; and so on [194,238,283].
The combination of metformin and dexamethasone synergistically eliminates MM cells by reducing AKT-mTOR signaling and inhibits cell proliferation [191]. In animal models, metformin significantly synergized with ritonavir to reduce the tumor burden of multiple myeloma [53]. Concomitant treatments with metformin and bortezomib inhibit the effect of UPR on GRP78, impair autophagosome formation and increase apoptosis [194].
Several studies have shown that being overweight, having high insulin levels and a history of diabetes are poor prognostic indicators for patients with multiple myeloma [284]. A retrospective study of 1240 patients with MM evaluated the impact of diabetes mellitus (DM) and antidiabetic drugs on clinical outcomes in MM, and the data suggested that metformin is associated with reduced death in progressive MM and that it improves MM-related prognosis [192]. In a retrospective study of 3287 U.S. veterans with diabetes diagnosed with MGUS, metformin users demonstrated the potential for a lower risk of progression relative to multiple myeloma than non-users (hazard ratio (HR) 0.47, 95%CI 0.25-0.87) [189]. These findings were extended by a case control study by Boursi et al. Using information from a U.K. database, Boursi et al. found that antidiabetic drugs may be associated with a reduced risk of active MM in patients with MGUS diabetes and with a reduced but not significant risk of MM among patients with diabetes exposed to metformin (OR 0.39, 95%CI 0.14-1.13), with the effect being observed mainly in patients who had been on treatment for at least 2 years, rather than those who had been on treatment for a shorter period of time [190]. Of course, there are studies that have come to different conclusions. One study of 2363 MGUS patients and 9193 controls showed no significant reductions in the risk of MGUS in patients with diabetes treated with metformin alone compared to those without diabetes (OR 0.77, 95%CI 0.56-1.05) [285].

Lymphoma
The incidence of lymphoma is currently increasing year by year [286]. In patients who have relapsed or are resistant to conventional chemotherapy, especially those with an aggressive subtype, the prognosis remains suboptimal. Shi et al. provided evidence of the ex vivo activity of metformin acting on human lymphoma cells, demonstrating that the metformin activation of AMPK inhibits the growth of B-and T-cell lymphomas via the inhibition of mTOR signaling, suggesting that metformin can be used in the treatment of lymphoma [193]. At the same time, metformin can enhance the tumor suppressive effect of common drugs for treatments such as asdriamycin, dacarbazine, venetoclax and rituximab [200,230,287].
Diffuse large B-cell lymphoma (DLBCL) is the most common histologic subtype of non-Hodgkin's lymphoma (NHL) [288]. The mechanisms by which metformin affects DLBCL cell proliferation are multifactorial. A portion of DLBCL survival and progression is highly dependent on B-cell receptor (BCR) signaling [289][290][291]. Metformin blocks cholesterol synthesis by inhibiting the key cholesterol synthesis-related enzyme hydroxy-3-methylglutaryl coenzyme A synthetase 1 (HMGCS1), which ultimately leads to the intracellular inhibition of cholesterol-dependent BCR and its downstream signaling, resulting in a blockage of BCR signaling [292]. Metformin acts on lymphoma cells to produce an increase in reactive oxygen species, and it promotes the loss of mitochondrial membrane permeability [200]. Metformin induced a G1 cell-cycle block and showed a decrease in cellular proteins C-MYC, cyclin dependent kinase 2 (CDK2), E2F and proliferating cell nuclear antigen (PCNA) [200]. Burkitt's lymphoma (BL) is a rare and aggressive subtype of NHL [293]. Metformin was found to strongly affect the viability of DAUDI cells (a human BL cell line), and peripheral blood lymphocytes (PBL) were unaffected by treatments, suggesting that metformin has the ability to selectively target cancer cells [228]. BL exhibits a specific metabolic shift towards aerobic glycolysis, leading to a strong change in metformin that can kill BL cells by determining glucose use [228].
The results of clinical studies on the role of metformin in lymphoma are mixed. A study by Wang et al. found no evidence of significant improvements in event-free (EFS), lymphoma-specific (LSS) and OS-associated metformin use in terms of prognostic improvement in newly diagnosed DLBCL and follicular lymphoma (FL) [294]. In addition, two large retrospective studies and one case-control study involving the action of metformin in patients with lymphoma showed similar results in that metformin exposure as a pre-chemotherapy agent had no significant effects on lymphoma survival [188,295,296].
There are other studies that disagreed with the above observations. A retrospective study on metformin action in patients with DLBCL in T2DM showed that the use of metformin did help significantly improve CR rates and ORR, in addition to PFS. In addition, the metformin group showed a trend toward improved OS, although the difference did not reach a statistically significant level [292]. A retrospective analysis of DLBCL patients receiving first-line chemotherapy showed that the use of metformin improved the PFS from 60 to 90 months and OS from 71 to 100 months in patients with diabetes compared with those not using metformin [200]. In a 10-year study in Taiwan on whether the use of metformin affects the risk of NHL, the risk of NHL was significantly lower in patients with T2DM treated with metformin-based antidiabetic drugs compared with non-metformin antidiabetic drugs [297]. One team also found significantly higher overall long-term survival for lymphoma in the metformin group (5.89 vs. 1.29 years, p < 0.001) than in the non-metformin group. Metformin was associated with a significant positive impact in improving overall survival in patients with type 2 diabetes and lymphoma [298]. In patients with DLBCL with DM, metformin improved the PFS (94 vs. 55.4 mon, p = 0.007) and OS (100 vs. 70.5 mon, p = 0.031) compared with other anti-diabetic agents [299]. A prospective study by Dr. Zhao's group also found that the use of metformin prolonged survival in DLBCL [300].

Ongoing Studies
According to ClinicalTrials.gov, there are currently 17 trials of metformin and HM. Metformin is being investigated in HM mainly as an adjuvant therapy. Five of these seventeen studies have been completed. These studies are shown specifically in Table 5.

Adverse Effects of Metformin
Metformin can cause various adverse reactions, which cannot be ignored in the treatment process. Digestive disturbances (nausea, vomiting, diarrhea, stomach upset, etc.) are the most common side effects of metformin and, while severe symptoms may lead to discontinuation in 5-10% of metformin users, these side effects are usually transient and mild [301][302][303]. This can be easily resolved by gradually titrating the dose, taking the drug with food or switching to an extended-release formulation [304]. The effects of metformin on gastrointestinal function may lead to vitamin B12 deficiency [305]. A cross-sectional study found a significant correlation between metformin ≥ 1500 mg/d and vitamin B12 deficiency [306]. This effect was more common in patients who had been on high doses for three years [307]. It can be reversed by discontinuation or calcium supplementation [308].
The most serious adverse effect is lactic acidosis, but this complication is very rare, with an incidence of 1/30,000, and it is mainly seen in patients with diabetes with renal and hepatic abnormalities [309]. Metformin increases plasma lactic acid levels by inhibiting mi-tochondrial respiration in the liver and inhibiting the conversion of lactic acid into glucose in the liver [310,311]. Therefore, excessive metformin can lead to a severe hypoglycemia. The true role of metformin in the development of metformin-associated lactic acidosis (MALA) in the clinical setting is unclear. Some trials have shown that the incidence of lactic acidosis in metformin users is not significantly different from that in patients treated with other oral anti-hyperglycemic effects agents [312]. The risk factors for MALA include impaired renal function resulting in poor metformin clearance, impaired liver function resulting in poor lactate clearance and increased lactate production, such as sepsis [313]. Overall, mortality in patients taking metformin or even developing lactic acidosis appears to be related to the underlying disease rather than metformin being a toxic drug [314]. As metformin is excreted by the kidneys, chronic kidney disease is considered a major risk factor for lactic acidosis [75,315]. There is insufficient evidence to support the safe use of metformin in people with eGFR values below 30 mL/min/1.73 m 2 . Therefore, it is generally recommended to discontinue metformin if renal function is below this level [316].
In-depth studies have identified several unexpected risks associated with metformin, such as birth defects in offspring, cognitive deficits in rodents, acute hepatitis, cholestasis and hemolytic anemia [317][318][319]. These require further confirmation of adverse effect results and revision of clinical guidelines for precisely defined indications and doses of metformin. A fact sheet issued to the public by the UK Teratology Information Service states that there are no known risks associated with the use of metformin during pregnancy [320]. However, there is limited evidence that metformin can affect anthropometric parameters in offspring with unknown long-term consequences. Therefore, the risks and benefits of continuing metformin during pregnancy should be carefully evaluated and individualized [321]. Metformin therapy appears to be safe for most patients in the early stages of COVID-19 infection. However, immediate discontinuation of metformin as soon as COVID-19 infection becomes severe is a reasonable and practical preventive measure for metformin. This approach may reduce the likelihood of life-threatening complications from COVID-19 or lactic acidosis events associated with metformin [322].

Limitations of Metformin in Cancer Treatment
Although some studies support the effectiveness and feasibility of metformin for the treatment of cancer, metformin is primarily used in patients with diabetes. In clinical use, the feasibility of metformin alone or in combination with other cancer treatment regimens may be limited to patients with non-diabetes because it may result in adverse effects that are intolerable to some patients [323]. For example, patients receiving both metformin and radiation therapy experienced increased local toxicity, increased frequency of treatment interruptions and desquamation or dermatitis [324][325][326]. The report also identified time-related biases in observational studies that tended to exaggerate metformin's anti-tumor effects [327,328]. The anti-tumor effects of metformin increased with increasing doses, as compiled from experimental data. Many experiments have used metformin at concentrations much higher than the conventional dose for diabetes treatment [329]. These high doses are not suitable for actual clinical use due to potential drug toxicity. At the same time, the efficacy of metformin as an anti-cancer drug is highly dependent on the glucose concentration in the tumor microenvironment [330,331].
In addition, many factors influence the effectiveness and responsiveness of metformin in cancer therapy. For example, few studies on metformin transporters, OCT, PMAT and MATE have provided insight into the accumulation of metformin within cancer cells to determine the specificity for cancer cells and whether normal non-cancerous cells are immune to these effects of metformin. The tissue expression of transporters mediating metformin uptake varies between normal and tumor cells and may be affected by various drugs. Poor uptake of metformin in target cells may limit its potential to treat cancer [90]. The heterogeneity of transporter expression in different cancers and how to obtain safe and effective doses of metformin in the treatment of cancer are areas of interest [332]. As with other chemotherapeutic treatment modalities, the long-term administration of metformin may be challenged by treatment-related drug resistance. Reports show that MCF7 breast cancer cells become cross-resistant to metformin and tamoxifen after long-term metformin therapy and are dependent on constitutive activation of Akt-Snail1-E-cadherin signaling [333]. Existing studies have shown inconsistent marker expression and survival outcomes for metformin as an anti-cancer agent in different settings. Variations in study design and potential biases, such as time affected by prior treatment and other relevant confounding factors, contributed to the variability of results [334].
Multiple factors influence the sensitivity of cancer cells to metformin, such as genetics, microenvironment, metabolic environment, biodistribution and tissue specificity [170]. It was demonstrated that the culture medium altered the sensitivity of cancer cells to metformin, as cells cultured in DMEM required up to 10 mM metformin to inhibit proliferation, whereas tumor cells cultured in an RPMI medium required a lower dose of metformin. At low glucose concentrations in cells lacking AMPK/LKB1, metformin is more sensitive to growth inhibition. Dietary restriction with intermittent fasting has been shown to enhance the response to metformin and metformin appears to impair tumor growth only when administered during fasting-induced hypoglycemia [335]. These are all issues that should be considered when treating cancer with metformin.

Conclusions
Metformin appears to be a new hope for HM treatment, and metformin acts as a pleiotropic compound that works together to prevent tumorigenesis via various pathways. The anti-cancer mechanism introduces the modulatory effects of metformin on multiple signaling pathways, including the activation of AMPK-dependent and AMPK-independent pathways, the improvement of cancer cell microenvironment, the amelioration of inflammatory responses and a reduction in tumor vascular metastasis. In addition, combining current anti-cancer drugs with metformin may improve their efficacy and reduce adverse drug reactions. Therefore, metformin can also be further investigated as an adjuvant therapeutic agent for HM. The field of oncology is among the most rigorous fields that use the gold standard of evidence in randomized controlled trials. Therefore, large, randomized, double-blind and placebo-controlled studies are necessary in order to find the best combination strategy by identifying the optimal dose of metformin, combination of drugs to avoid adverse events and potential biomarkers to implement this strategy in a precise oncological context. Ultimately, the efficacy of metformin in HM will be tested and the utility of the drug will be clarified.