Metformin: The Answer to Cancer in a Flower? Current Knowledge and Future Prospects of Metformin as an Anti-Cancer Agent in Breast Cancer

Interest has grown in studying the possible use of well-known anti-diabetic drugs as anti-cancer agents individually or in combination with, frequently used, chemotherapeutic agents and/or radiation, owing to the fact that diabetes heightens the risk, incidence, and rapid progression of cancers, including breast cancer, in an individual. In this regard, metformin (1, 1-dimethylbiguanide), well known as ‘Glucophage’ among diabetics, was reported to be cancer preventive while also being a potent anti-proliferative and anti-cancer agent. While meta-analysis studies reported a lower risk and incidence of breast cancer among diabetic individuals on a metformin treatment regimen, several in vitro, pre-clinical, and clinical studies reported the efficacy of using metformin individually as an anti-cancer/anti-tumor agent or in combination with chemotherapeutic drugs or radiation in the treatment of different forms of breast cancer. However, unanswered questions remain with regards to areas such as cancer treatment specific therapeutic dosing of metformin, specificity to cancer cells at high concentrations, resistance to metformin therapy, efficacy of combinatory therapeutic approaches, post-therapeutic relapse of the disease, and efficacy in cancer prevention in non-diabetic individuals. In the current article, we discuss the biology of metformin and its molecular mechanism of action, the existing cellular, pre-clinical, and clinical studies that have tested the anti-tumor potential of metformin as a potential anti-cancer/anti-tumor agent in breast cancer therapy, and outline the future prospects and directions for a better understanding and re-purposing of metformin as an anti-cancer drug in the treatment of breast cancer.


Background and Introduction
The reinforced link between diabetes and cancer and/or breast cancer has generated interest in studying the effects of widely prescribed anti-hyperglycemic/anti-diabetic drugs on the risk, incidence, progression, response to therapy, resistance, and post-therapeutic relapse in cancers. The incidence and rising rates of diabetes is a serious concern in the medical field around the globe. Worldwide, the World Health Organization (WHO) data estimate nearly 422 million diabetes affected individuals in 2014, a sharp rise when compared to the 108 million diabetes affected individuals in 1980 [1]. Data projections suggest that the global diabetes prevalence of 8.8% in 2017 will further increase to 9.9% Interest has grown in studying the possible use of metformin as an anti-cancer/anti-tumor agent individually or in combination with frequently used chemotherapeutic agents and/or radiation. Epidemiological studies and meta-analysis data suggest that diabetic individuals on a metformin treatment regimen, to control their blood glucose levels, have a lower risk of developing cancers of all types and additionally individuals who are both diabetic and suffering from cancer and on metformin treatment have an improved response to chemotherapy and radiation therapy, better prognosis, and higher survival rates when compared to those who do not take metformin [5,[35][36][37][38][39][40][41][42]. In cancer cells, the ability of metformin to alter cancer metabolism and mitochondrial function and Figure 1. Multifaceted benefits of metformin: Metformin reduces blood glucose levels in circulation by decreasing gluconeogenesis and glycogenolysis in the liver, decreasing the intestinal absorption of glucose, reducing the release of free fatty acids (FFA) from adipose tissue, and increasing glucose utilization by the muscle. Metformin exerts its cardioprotective effects by increasing cardiac FFA oxidation and glycolysis, reducing ischemia-associated stunning and infarct size, decreasing cardiac hypertrophy, apoptosis, and fibrosis, thereby improving cardiac functions (systolic and diastolic). Metformin's vasculo-protective effect is accounted for by its effect on reducing inflammation, endothelial apoptosis, oxidative stress, and fibrosis of the vasculature, improving both endothelial and smooth muscle cell function and inhibiting epithelial mesenchymal transition (EMT) transition, thus curbing vascular remodeling and causing overall improvement of vascular function. In addition, metformin exerts its anti-cancer effects by decreasing incidence of different cancers and inhibition of proliferation and migration of cancer cells, activation of apoptosis, and reducing EMT and metastasis.
Several epidemiological and meta-analysis data and in vitro, pre-clinical, and clinical studies also link the beneficial outcomes in the treatment of different forms of breast cancer to metformin treatment either individually or in combination with chemotherapy and radiation therapy. In the current article, we discuss the biology of metformin and its molecular mechanism of action, the existing cellular, pre-clinical, and clinical studies that have tested the anti-tumor potential of metformin as a potential anti-cancer/anti-tumor agent in breast cancer therapy and outline the future prospects and directions for a better understanding and re-purposing of metformin as an anti-cancer drug in the treatment of Biomolecules 2019, 9,846 4 of 33 breast cancer. We aim to encourage the scientists working with metformin in breast cancer to address unanswered questions pertaining to areas such as breast cancer specific therapeutic dosing of metformin, specificity of metformin to breast cancer cells at high concentrations, resistance of specific forms of breast cancer to metformin therapy, the efficacy of combinatory therapeutic approaches, post-therapeutic relapse of the disease, and efficacy in breast cancer prevention in non-diabetic individuals.

Biology of Metformin and Molecular Mechanism of Action
In a type 2 diabetic individual who receives metformin orally, the concentration of metformin in the hepatic circulation may reach 50 µM; with the peak plasma concentration of metformin at 20 µM [5,48,49]. The hydrophilic and cationic nature of metformin at physiological pH makes it highly unlikely that metformin rapidly diffuses through the cell membrane and exerts it effect on cell function. In addition, the kidneys carry out the elimination of unaltered metformin through the urine [50]. Hence, it is evident that metformin requires the presence and support of transporter molecules for its absorption, distribution, and elimination to exert its biological function. In this regard, the organic cation transporters 1, 2, and 3 (OCT1, OCT2, and OCT3), the plasma membrane monoamine transporter (PMAT), and multidrug and toxin extrusion protein 1 and 2 (MATE1 and MATE2) transporters are reported to play key roles in transporting metformin into and out of the cell in the intestine, liver, and kidney [50][51][52][53][54][55][56][57]. The thiamine transporter 2 (THTR2) also plays a role in intestinal absorption and renal re-absorption of metformin [58]. Alterations in the OCT1 gene reduced hepatic uptake of metformin and reduced the efficacy of metformin in reducing blood glucose levels by the inhibiting gluconeogenesis and glycogenolysis [59,60].
While several studies have reported various 'AMPK dependent' and 'AMPK independent' mechanisms for the anti-cancer/anti-tumor effects of metformin in cancer therapy, these anti-cancer effects of metformin were only observed at very high concentrations (>5 mM) and fall short of explaining how such high concentrations enters the cancer cells and exerts its anti-neoplastic effect. Studies have implicated that the susceptibility and/or resistance of cancer cells to metformin treatment is dependent on the varying levels of the cell surface metformin transporters. Overexpression of OCTs that contribute to intracellular accumulation of metformin in cancers would make them susceptible to metformin treatment (which should explain the high concentrations of metformin required for anti-cancer treatment) while the overexpression of MATE transporters that contribute to the extrusion of metformin out of the cell would render the cancer cell resistant to metformin treatment ( Figure 2) [61][62][63][64].
There are two general mechanisms that could explain the putative anti-cancer effects of metformin. The 'indirect' anti-cancer effects of metformin arise from its ability to reduce insulin resistance, insulin levels, and fasting glucose levels [65]. Physiologically, insulin and insulin-like growth factor-1 (IGF1) largely regulate carbohydrate and lipid metabolism and storage and protein synthesis via transmembrane receptor binding and activation of receptor tyrosine kinase and subsequent activation of intracellular insulin receptor substrate-1 (IRS1); however, insulin and IGF1-mediated signaling pathways are also implicated in pathogenesis and progression of several cancers via the activation of the Ras/Raf/MEK/ERK, PI3K/Akt/mTORC1, and GSK3β/β-catenin pathways [5,44]. Metformin reduces blood glucose levels in circulation by decreasing gluconeogenesis and glycogenolysis in the liver, decreasing the intestinal absorption of glucose, reducing the release of FFA from adipose tissue, and increasing glucose utilization by the muscle. Lower blood glucose concentration in turn decreases the synthesis and secretion of insulin by the β-cells of the pancreas and reduces the levels of insulin in circulation. The anti-proliferative activity of metformin in several cancers is at least in part attributed to its ability to reduce the levels of insulin/IGF1, which in turn inhibits the insulin/IGF1 mediated molecular pathways that support tumor initiation and progression rather than a direct anti-proliferative/anti-cancer action ( Figure 2) [44].
Metformin also exhibited 'direct' anti-cancer effects in many different cancer related studies. Since cancer cells are known to utilize glucose rapidly through glycolysis (Warburg effect) to meet their energy needs when compared to normal cells, metformin-mediated decrease in glucose levels should also curb tumor growth, although reports suggest that cancer cells use alternative sources of energy when starved of glucose or when glycolysis is inhibited [45,[66][67][68]. Furthermore, AMPK inhibits IRS1 mediated IR and IGFR oncogenic signaling via PI3K/Akt/mTOR, which, potentially, also contributes to the anti-cancer effect of metformin ( Figure 2) [69][70][71]. and reduces the levels of insulin in circulation. The anti-proliferative activity of metformin in several cancers is at least in part attributed to its ability to reduce the levels of insulin/IGF1, which in turn inhibits the insulin/IGF1 mediated molecular pathways that support tumor initiation and progression rather than a direct anti-proliferative/anti-cancer action ( Figure 2) [44].
Metformin also exhibited 'direct' anti-cancer effects in many different cancer related studies. Since cancer cells are known to utilize glucose rapidly through glycolysis (Warburg effect) to meet their energy needs when compared to normal cells, metformin-mediated decrease in glucose levels should also curb tumor growth, although reports suggest that cancer cells use alternative sources of energy when starved of glucose or when glycolysis is inhibited [45,[66][67][68]. Furthermore, AMPK inhibits IRS1 mediated IR and IGFR oncogenic signaling via PI3K/Akt/mTOR, which, potentially, also contributes to the anti-cancer effect of metformin ( Figure 2) [69][70][71]. Figure 2. Cellular anti-cancer/anti-tumor effects of metformin: The hydrophilic and cationic metformin is transported into the cell via the organic cation transporters (OCT), which support the intracellular accumulation of metformin. The anti-proliferative activity of metformin in several cancers is at least in part attributed to its ability to reduce the levels of insulin/IGF1, which in turn inhibits the insulin/IGF1 mediated molecular pathways that support tumor initiation and progression. Metformin treatment directly activates AMPK and the 'AMPK dependent' effects The hydrophilic and cationic metformin is transported into the cell via the organic cation transporters (OCT), which support the intracellular accumulation of metformin. The anti-proliferative activity of metformin in several cancers is at least in part attributed to its ability to reduce the levels of insulin/IGF1, which in turn inhibits the insulin/IGF1 mediated molecular pathways that support tumor initiation and progression. Metformin treatment directly activates AMPK and the 'AMPK dependent' effects include inhibition of c-Myc, NF-κB, and mammalian target of rapamycin-C1 (mTORC1) pathways and acetyl Co-A carboxylase (ACC)-dependent lipogenesis pathways while activating the p53 pathway and DICER-mediated miRNA synthesis. Metformin, albeit at high concentrations, is also known to inhibit the mitochondrial Complex 1 of the electron transport chain (ETC) thereby reducing ATP, levels which increases the AMP/ATP ratio further leading to AMPK activation. A decrease in ATP/energy levels can also lead to mismanaged protein folding mechanisms leading to the accumulation of unfolded or misfolded proteins and prolonged unfolded protein response (UPR) without rectification of endoplasmic stress triggers apoptosis through multiple mechanisms, which include activation of UPR mediated apoptotic/death signaling and activation of autophagy and subsequent autophagic cell death. AMPK independent metformin treatment-associated anti-cancer effects are mediated by Rag GTPases, REDD1, and STAT3. Overall metformin treatment in cancer cells causes translational inactivation, cell-cycle arrest, inactivation of cell proliferation and migration, inhibition of UPR, promotes apoptosis, inhibits EMT, invasion, and metastasis, and increases susceptibility to chemotherapy/radiation therapy.
Several experimental studies reported the 'direct' anti-cancer effects of metformin, which are distinct from its 'indirect' anti-cancer effects that are related to its anti-hyperglycemic actions, inhibition of hepatic gluconeogenesis, and reduction of insulin signaling [72][73][74]. In cancer cells, aberrant signaling mechanisms were reported involving key proteins and their related pathways associated with protein synthesis and survival (mTOR, c-Myc, and NF-κB), lipid synthesis (ACC), DNA damage repair and apoptosis (p53), and miRNA synthesis and function (DICER). Each of these proteins and their modulation/regulation can impact the incidence, growth, and progression of malignant tumors. Mammalian target of rapamycin-C1 (mTORC1) is activated and upregulated by nutrients, growth factors and energy and stress signals, and key signaling pathways (PI3K, MAPK, and AMPK) and is inhibited by rapamycin [75][76][77]. Activation of mTORC1 regulates cellular protein synthesis and cell survival through the phosphorylation of its substrates, 4EBP1, and p70S6K [75][76][77]. The central role of mTOR in regulating cellular protein synthesis and cell survival explains the association of an overactive mTOR pathway cancer [75][76][77]. Several mTOR inhibitors were successfully tested for the treatment of various cancers [75][76][77]. Cellular Myc (c-Myc) is a well-studied oncogene which was constitutively overexpressed in various cancers [78,79]. Myc reportedly regulates an array of cellular functions that support cancer cell growth and progression including transcription of several oncogenes, translation, cell cycle progression, cell proliferation, differentiation and survival, ribosome biogenesis, signal transduction, and also cancer stem cell-related signaling and resistance to cancer treatment [78][79][80]. NF-κB activation, similarly, promotes cancer growth and by regulating transcription of genes that support cell proliferation and survival, angiogenesis, tumor progression, and metastasis [81][82][83]. The DNA repair and pro-apoptotic nuclear transcription factor, p53, is vital to tumor suppression [84]. Loss of p53 function and mutations in the p53 gene are notably the cause for the incidence and progression of many different cancers by supporting cell proliferation and survival, metabolism, genome instability, pro-survival autophagy, and metastasis in addition to conferring therapeutic resistance to cancer cells [84][85][86]. The RNase enzyme, DICER, that is important in processing the formation of function micro-RNAs (miRNA) was frequently downregulated in human cancers and was linked to cancer progression and cancer metastasis [87][88][89].
Furthermore, metformin can inhibit the mitochondrial respiratory chain complex 1 (Figure 2), thereby causing a reduction in the NADH oxidation and the proton gradient across the inner mitochondrial membrane subsequently reducing the rate of oxygen consumption [113,114]. Since cancer cells are highly glycolytic in nature (Warburg effect) and depend less on the oxidative phosphorylation for its energy needs (ATP), it can be argued that the effect of metformin as an inhibitor of the electron transport chain (ETC) complex 1 may be weak and reversible and may not impact the growth or proliferation of cancer cells [47]. However, any cancer cell that utilizes oxygen for mitochondrial respiration would produce mitochondrial ATP and, thus, a decrease in ATP production due to metformin-mediated inhibition of ETC complex 1 should be toxic to the cells [47]. Additionally, the increasing levels of AMP due to ETC complex 1 inhibition should also in turn activate AMPK ( Figure 2) [46,115]. The decrease in ATP also contributes to the accumulation of unfolded/misfolded proteins [116][117][118]. The accumulation of misfolded or unfolded proteins turns on the endoplasmic reticulum stress/unfolded protein response (UPR) pathway [119]. Prolonged UPR and accumulation of unfolded proteins without rectification of endoplasmic reticulum stress triggers apoptosis through multiple mechanisms, which include activation of UPR-mediated apoptotic/death signaling and activation of autophagy and subsequent autophagic cell death [119]. Additionally, the ER stress mediated release of calcium (Ca 2+ ) from the endoplasmic reticulum stores leads to Ca 2+ accumulation in the mitochondria causing depolarization of permeability transition pore (PTP) and inducing apoptosis via the release of caspases [119].
Metformin treatment-associated 'AMPK independent' anti-cancer effects are mediated by regulated in DNA Damage-1 (REDD1; also known as DNA damage inducible transcript-4-DDIT4), Rag GTPases, and signal transducer and activator of transcription-3 (STAT3) (Figure 2). REDD1/DDIT4 is known to be inhibitor of mTOR signaling and thereby possess tumor suppressive properties by inhibition of protein synthesis and cell survival [120][121][122][123]. Metformin reportedly activated the p53/REDD1 axis to cause AMPK independent inhibition of mTOR in cancer cells ( Figure 2) [124]. Activation of the p53/REDD1 pathway also reduces the expression of Cyclin D1, thereby reducing cell proliferation [125]. Rag GTPases, a sub-family of Ras-related GTPases, is involved in amino acid signaling mediated activation and functioning of the mTOR pathway [126][127][128]. Metformin treatment inhibits the mTOR pathway via the inhibition of Rag GTPases in cancer cells, which in turn reduces protein synthesis and causes cell cycle arrest ( Figure 2) [125,129,130]. The aberrant activity of STAT3 has been implicated in promoting the pro-oncogenic functions such as initiation, progression, metastasis, and immune evasion in different cancers [131,132]. Overexpression of STAT3 contributes to cell survival, proliferation, cell cycle progression, anti-apoptosis, migration, invasion, angiogenesis, chemoresistance, immunosuppression, and self-renewal and differentiation of stem cells by regulating the expression of its downstream target genes [131,132]. Metformin treatment inhibited STAT3 nuclear translocation and exerted anti-proliferative, anti-metastatic, and pro-apoptotic effects in cholangiocarcinoma cells and breast cancer cells ( Figure 2) [133,134].
The multifaceted ability of metformin to influence cancer cell growth and cancer progression through various molecular mechanisms, as discussed above, has made it an interesting candidate drug with potential in the treatment of breast cancer. In the following sections of the article, we briefly discuss the cellular, pre-clinical, and clinical studies that are currently testing metformin as a monotherapy or in combination with other chemotherapeutic drugs or phytochemicals/natural compounds for its efficacy as an anti-cancer/anti-tumor agent in the treatment of estrogen receptor (ER) positive, progesterone receptor (PR) positive, human epidermal growth factor receptor 2 (HER2) positive types of breast cancers, and triple negative breast cancers (TNBCs) [135].

Cellular and Pre-Clinical Data
The report by Evans JM, et al. in 2005 suggesting that metformin-treated diabetic patients had a reduced risk of cancers spurred a massive interest in studying the anti-cancer/anti-tumor effect of metformin giving rise to several in vitro and pre-clincal studies [39]. As elaborated in our article on the link between diabetes and breast cancer, insulin and IGFs promote the incidence and progression of breast cancer through several mechanisms that support translational activation, cell cycle progression, activation of cell proliferation and migration, inhibition of apoptosis, activation of EMT, increasing invasiveness and metastasis, and conferring resistance to chemotherapy [5]. The ability of metformin to promote glucose uptake by the muscles, increase insulin senitivity, reduce insulin levels, and thereby reduce blood glucose levels in itself reduces pre-neoplastic and neoplastic breast cell proliferation [136]. The fact that metformin has remained off patent since 2002, is easy to synthesize and economical, well-tolerated with very little side effects, and works at the level of the whole organism by reducing insulin levels and also directly on the tumor cells makes metformin an interesting drug of choice for the treatment of cancer [137].
Althought several of the studies reported metformin treatment-associated inhibition of cancer cell growth and proilferation, activation of cancer cell death, inhibition of invasion and metstasis, and tumor regression, these beneficial effects of metformin were observed at only significantly high concentrations (>5 mM), at least 100-fold higher than with the peak plasma concentration of metformin when administered orally for the treatment of type 2 diabetes [48,49]. As discussed earlier, the selective accumulation of metformin and therefore the sensitivity of cancer cells to metformin should depend on the levels of expression of the various transporters (OCTs, PMAT, and MATEs) in the cells [50][51][52][53][54][55][56][57]59,60].
In estrogen receptor (ER) and progesterone receptor (PR) positive MCF7 cells and triple negative breast cancer cells (BT20 and MDA-MB-468), overexpression of OCT3 was related to enhanced metformin uptake and anti-tumor efficacy, indicating that OCT3 supported the movement of metformin into the cancer cell [61]. In a rat model of 1-methyl-1-nitrosourea initiated mammary tumor, the intracellular accumulation of metformin, activation of AMPK, decrease in tumor volume and proliferation, and subsequent suppression of tumor progression correlated with the higher expression of OCT2, which also transported metformin into the cancer cell [62,64]. On the other hand, a study involving 19 different cancer cell lines identified that high MATE2 (which transports metformin out of the cell) expression levels correlated with the cancer cell resistance to the anti-proliferative effect of metformin [63].
The efficacy of metformin as an anti-cancer agent may also depend on the blood glucose levels in the treated individual. It was reported that in a normoglycemic condition, metformin treatment protected normal cells while causing cell cycle arrest in cancer cells of the breast [138]. Metformin treatment had little effect on breast cancer cells grown in hyperglycemic conditions, which promoted growth and aggressiveness of the cells [139]. On the other hand, metformin treatment in glucose-starved/deprived breast cancer cells induced cell death [139]. We have shown that the ability of metformin to inhibit the Akt/mTOR pathway, induce G 2 /M cell cycle arrest, and cell death in microvascular endothelial cells overexpressing VEGF, was enhanced under glucose-starved conditions [68,138]. This was shown to be true in other cancer cells as well [68,138]. In MCF7, SKBR3, and MDA-MB-231 cells, metformin effectively activated AMPK dependent apoptosis (partially independent of mTORC1) at physiological glucose conditions while the efficacy of metformin was lost under high glucose or amino acid rich conditions [140]. In fact, under nutrient poor conditions, metformin shifted the cellular glycolytic equilibrium through the AMPK dependent downregulation of pyruvate kinase M2 [140]. Under high glucose conditions, the energy derived from aerobic glycolysis promoted cell proliferation, conferred resistance to metformin treatment, and protected the triple negative breast cancer cells (TNBCs) from metformin induced apoptosis [141].
Metformin treatment in MCF7 cells reportedly caused a 30% reduction in global protein synthesis, which was associated with the AMPK dependent inhibition of the mTOR pathway [107]. In MDA-MB-231 cells lacking LKB1, an upstream kinase that phosphorylates AMPK, and in TSC2 null embryonic fibroblasts of mouse origin, treatment with metformin did not have any effect, indicative of the fact that the mechanism of action of metformin should be dependent on LKB1 and TSC2 [107]. In MCF7 cells, treatment with metformin induced oxidative stress, AMPK and FOXO3a mediated cell death and cell cycle arrest [142]. In breast carcinoma cells, metformin treatment modulated the expression and function of tumor suppressor p53 and reduced the levels of cyclin-D1 thereby causing cell cycle arrest and inhibiting tumor cell growth [142,143]. In TNBCs, treatment with metformin increased the levels phosphorylated-AMPK and induced PARP cleavage in a dose-and time-dependent manner, reduced the levels of phosphorylated EGFR, total EGFR, phosphorylated-MAPK, phosphorylated-Src, cyclin D1, and cyclin E, thus causing and inhibition cell proliferation (partial S phase arrest), colony formation, and promoting apoptosis [144]. In nude mice bearing tumor xenografts of the MDA-MB-231 TNBC cells, metformin treatment significantly reduced tumor growth and cell proliferation when compared to untreated controls [144]. Furthermore, a significant decrease in tumor growth and occurrence was observed in the nude mice, which were pre-treated with metformin prior to the administration of MDA-MB-231 cells [144]. Metformin treatment also inhibited several key enzymes associated with glucose metabolism in MDA-MB-468 TNBC cells, thus indicating that metformin maybe efficient in the treatment of TNBCs [145]. In an in vivo model of metastatic breast cancer, metformin treatment mediated downregulation of platelet-derived growth factor B (PDGF-B) inhibited angiogenesis and the formation of immature vasculature and thus contributed the inhibition of metastasis and further sensitized the metastatic breast cancers to chemotherapy [146].
In MCF7 breast cancer cells, uncoupled reactions accounted for a major fraction of the cellular respiration [147]. Metformin treatment caused a dose-dependent decrease in mitochondrial respiration in the MCF7 cells and led to an upregulation of glycolysis and reduction in cell proliferation [147].
While metformin treatment was associated with inhibition of cell growth in six different basal cancer cells (MDA-MB-468, HCC70, HCC1806, MDA-MB-231, BT20, and HCC1937), the MDA-MB-468 and HCC70 cells were most sensitive [134]. Additionally, a significant reduction in tyrosine and serine phosphorylation of STAT3 was observed in metformin-treated MDA-MB-468, HCC70, MDA-MB-231, and BT20 cells [134]. While overexpression of constitutively active STAT3 negatively affected the efficacy of metformin in these cells, knocking down STAT3 enhanced metformin induced apoptosis [134]. Sequential modulation of DICER and c-Myc played a critical role in metformin mediated anti-cancer effects in MCF7 and BT474 breast cancer cells [105]. In fact, metformin modulated the c-Myc levels in breast cancer cells through its action on miRNA33a, an effect which was reportedly abolished in DICER knockdown SUM159PT cells [105].

Clinical Data and Trials
Diabetic subjects on metformin were reported to have a significantly lower risk of developing breast cancer when compared to diabetic individuals who were not using metformin to control their blood glucose levels [148,149]. A lower incidence of invasive breast cancer was observed in diabetic patients taking metformin when compared to those on other anti-diabetic (such as sulfonylureas) treatment plans [150]. This correlation is, however, independent of diabetes and thus supports the use of metformin in non-diabetic individuals for cancer prevention. On the other hand, high levels of circulating insulin as well using insulin and insulin analogues to control blood glucose levels were associated to an increased risk of breast cancer [149,151]. The decrease in blood glucose and insulin levels as a result of metformin administration reduced the proliferation of cancer cells and suppressed the growth of the tumor [149,151]. The metformin treatment-associated decrease in circulating hormone (estrogen) levels further supports a reduction in breast cancer incidence among metformin-administered diabetic subjects [149,152,153]. Since breast cancer cells is addicted to glucose utilization via aerobic glycolysis for growth and proliferation, metformin treatment-related increase in energy stress and decrease in glucose levels should also contribute to the suppression of breast cancer growth [149]. Additionally, the reduction in the mitochondrial ATP production as a result of complex I inhibition by metformin further makes the cancer cells susceptible to metformin treatment [47].
A large study involving 2529 breast cancer patients showed a 'complete response', with no traces of cancerous cells in the affected breasts or lymph nodes, in diabetic breast cancer patients receiving a combination of chemotherapy and metformin [154]. Additionally, an improved pathologic complete response rate was observed in diabetic breast cancer patients who received both metformin and neoadjuvant chemotherapy than the diabetic patients who received only chemotherapy [154]. The phase 3 adjuvant lapatinib and/or trastuzumab treatment optimization (ALTTO) randomized trial reported that, while insulin administration supported cancer progression and was associated with detrimental effects, metformin-treated HER2 positive diabetic breast cancer patients presented with improved disease-free survival and overall survival rates when compared to HER2 positive diabetic breast cancer patients not on metformin treatment [155]. Although no significant association between exposure to metformin and the occurrence of breast cancer was reported, reports suggest that metformin may improve survival in breast cancer patients with diabetes since metformin administration to breast cancer subjects with diabetes reportedly led to a 45% risk reduction for all-cause mortality [156].
Data from ClincalTrials.gov and cancer.gov showed a list of 44 (search performed on 11 November 2019; Table 1) metformin administration-based clinical trials in breast cancer. While 12 out of the 44 metformin and breast cancer related clinical trials were terminated, withdrawn, or the status remains unknown, 14 clinical trials (out of the 44) were completed, 11 (of the completed 14) of which are specifically related to breast cancer (Table 2), while 3 (of the completed 14) include studies related to breast cancer and other solid tumors of the lung, kidney, liver, and endometrium. From the 11 completed metformin and breast cancer specific clinical trials, only 6 have reported the data or published articles related to the clinical trial. Eighteen breast cancer related clinical trials are currently active, out of which 14 are active and recruiting patients while 4 clinical trials have not started recruiting subjects for the study. Metformin is used as the sole or one of the interventions in nine (Table 2) out of the 14 active and recruiting clinical trials that are specific for breast cancer.
A higher incidence of post-menopausal breast cancer and deteriorating prognosis was observed in obese women when compared to post-menopausal women with normal body mass index (BMI) [157]. Both insulin resistance and the levels of estrogens (both endogenous and exogenous) contribute largely to the effect of obesity on breast cancer risk [5,157]. Exemestane, an irreversible aromatase inhibitor, decreases estrogen levels and is used routinely in the treatment of hormone receptor positive breast cancer [157]. While metformin reduces blood glucose levels by decreasing hepatic gluconeogenesis and glycogenolysis, increasing utilization of glucose of muscles, and improves insulin sensitivity, rosiglitazone improves insulin sensitivity via the activation of the PPARγ receptors [157]. A phase I dose escalation study (NCT00933309, Table 2) assessing the tolerability and pharmacokinetics of exemestane (25 mg/day, orally) in combination with metformin (dose level 1: 1500 mg/day, dose level 2: 2000 mg/day) and rosiglitazone (dose level 1: 6 mg/day, dose level 2: 8 mg/day) in post-menopausal overweight or obese women with hormone receptor positive metastatic breast cancer reported that the oral administration of exemestane with a combination of metformin and rosiglitazone was well tolerated [157]. Administration of metformin (1500 mg, daily, NCT00930579, phase II, Table 2) to 35 non-diabetic overweight women (BMI ≥ 25 kg/m 2 ) with breast cancer (stages 0-III) showed no significant difference in the tumor proliferation index (Ki67) when compared to age, BMI, and stage matched historic controls, despite a significant reduction in BMI, cholesterol, and leptin levels in the metformin-treated subjects [158]. Increase in the levels of Raptor, C-Raf, Cyclin B1, Cyclin D1, TRFC, and Syk; while reduction in the levels of pMAPK pT202, Y204 , JNK pT183, pT185 , Bad pS112 , PKCα pS657 , and Src pY416 was observed in the metformin administered (1500 mg, daily, NCT00930579, phase II, Table 2) non-diabetic overweight women with breast cancer when compared to age, BMI, and stage-matched historic controls [159]. This is indicative of the fact that in a clinical setting, metformin administration can influence cancer cell apoptosis, cell cycle, cell signaling, and invasion [159]. A phase II, single arm, 'window of opportunity' neoadjuvant metformin clinical study (NCT00897884, phase II, Table 2) conducted in non-diabetic early stage breast cancer patients investigated whether taking metformin (500 mg; thrice daily for ≥ 2 weeks after diagnostic biopsy) until surgery could reduce cell proliferation rates in the tumor tissue and reported reduced levels of phosphorylated Akt and ERK1/2, coupled to reduction on the levels of insulin and insulin receptors indicating that the anti-cancer effect of metformin presents insulin-dependent effects in a clinical scenario [136]. Interestingly, the levels of phospho-AMPK and phospho-ACC levels decreased in post-metformin treated breast tissues [136]. The metformin transporter OCT1 was found to be expressed in all the tissues examined [136]. A clinical trial (NCT01266486, phase II, Table 2) integrated dynamic positron emission tomography, metabolomics, and transcriptomics in breast cancer patients to study the effects of metformin treatment (500 mg for days 1-3, 1000 mg for days 4-6, and 1500 mg thereafter, minimum of 13 days and maximum of 21 days) and reported that metformin administration increases 2-deoxy-2-( 18 F)-fluoro-d-glucose (18FDG) influx into the tumors and activates genes related to mitochondrial metabolism [160]. Although no significant correlation was observed between baseline OCT1 levels and tumor metformin levels, it was interesting that the patient with highest tumor metformin levels correlated to the overexpression of tumor OCT1 [160]. A higher proliferation score was observed in tumors with upregulated genes related to oxidative phosphorylation [160]. The anti-tumor effect of metformin in breast cancer affects breast cancer metabolism and links these anti-cancer effects of metformin to mitochondrial metabolism thereby affecting cell proliferation at a transcriptional level [160]. A phase II clinical trial (NCT01589367, phase II, Table 2), compared the efficacy of a combination of letrozole (2.5 mg/day) and metformin (1000 mg/day for week 1, 1500 mg/day for week 2, 2000 mg/day for week 3) vs. letrozole (2.5 mg/day) and placebo, in ER-positive breast cancer patients [161]. A phase II (NCT01310231, phase II, Table 2) trial showed no significant association between the use of metformin and the stage of breast cancer, characteristics of the tumor at diagnosis, and survival of patients [162,163].

Monotherapy vs. Combination Therapy
The use of surgery, chemotherapy, and radiation therapy have become the norm for the treatment for most cancers with considerable success in suppressing tumor progression and improving the survival and quality of life of the affected individuals [164]. The hurdle, however, lies in the fact that the use of a single treatment modality/monotherapy approach leads to toxicity and the evolution of resistance, metastasis, and relapse of the disease. Much of the research in the field of oncology is now focused on overcoming these obstacles by treating cancers with a combination of one or more drugs or therapeutic modalities, which was found to be advantageous over monotherapies due to a number of reasons [137,164]. Using targeted therapeutic combination regimens translates to (1) higher efficacy of the treatment and improved overall outcome and prognosis, with decreased incidence of metastasis and relapse; (2) synergistic effects of the drugs, therefore requiring combination drugs to be used at much lower dosage but with maximum efficiency; (3) lesser drug induced toxicity and adverse side effects as a result of lower dosage of the drugs when used in combination; and (4) avoidance of resistance development against the drugs when used in combination [165]. A sound understanding regarding drug interactions and contraindications is necessary to formulate an appropriate drug combination to achieve therapeutic efficiency.
Combinations of two or more chemotherapeutic drug are currently being clinically tested for their efficacy in the treatment of various cancers with satisfactory outcomes [164]. Interest has peaked in combination therapeutic regimens, which employ a chemotherapeutic drug (hormone modulating drugs, anti-metabolite drugs, antibiotics, drugs targeting the structure and function of DNA, drugs that modulate protein translation) or natural compounds/phytochemicals in combination with metformin since it targets cancer cell metabolism and causes energy stress with minimal toxicity, both as a neoadjuvant and adjuvant therapeutic modalities. The feasibility of using of using metformin as one of the drugs in the combination is credited to the multifaceted anti-cancer/anti-tumor ability of metformin in targeting different molecular mechanisms that disrupt the growth and progression of cancers [164]. Several studies have used combination therapy approaches using metformin and phytochemicals/other chemotherapeutic drugs, which have been provided in Table 3. Table 3. Metformin-based combinatorial therapy in the treatment of breast cancers.

Conventional Chemotherapeutic Drug or Treatment Modality Effect and Possible Mechanism of Action Cells/in vivo Model Used
Ref.
MCF7/DOX resistant cells [170] 5-Fluorouracil, Epirubicin and Cyclophosphamide (FEC) Metformin synergizes FEC combination therapy via AMPK dependent mechanism in non-stem/parental breast cancer cells, while in cancer stem cells (CSCs) the synergistic effect of the combination treatment was found to be independent of AMPK. In CSCs, while metformin accelerated glucose consumption and lactate production, the production of intracellular ATP was significantly diminished leading to energy stress and impairment of the ability of CSCs to repair the FEC induced DNA damage. 4T1, BALB/c mice inoculated with 4T1 cells [172] Metformin treatment alone altered morphology decreased viability and migration of ER+ve MCF7 cells. The combination of aspirin and metformin synergistically altered morphology decreased viability and migration in TNBC MDA-MB-231 cells. HER2+ve SK-BR-3 cells showed a partial response to monotherapy (aspirin or metformin) and combinatorial therapy (aspirin and metformin).
MCF7, SK-BR-3, MDA-MB-231, [173] Chrysin Synergistic growth inhibitory effects due to suppression of hTERT and cyclin D1 gene expression T47D [174] Curcumin Inhibition of tumor proliferation and growth associated with reduced VEGF expression and angiogenesis, induction of p53 independent apoptosis, and activation of Th2 related immune response with no toxicity. EMT6/P cells, BALB/c mice inoculated with EMT6/P cells [175] Combination of PEGylated PLGA nanoparticle co-encapsulated metformin and curcumin exhibited dosage dependent toxicity and synergistic antiproliferative effect causing significant cell cycle/growth arrest in the cancer cells. The hTERT gene expression was significantly inhibited in cells treated with the nano-formulation of metformin-curcumin when compared to delivery of either metformin or curcumin alone.
T47D [176] Denosumab BRCA1 haplo-insufficiency driven RANKL gene overexpression was hampered by metformin treatment and disrupted the RANKL mediated auto-regulatory feedback in CSCs thereby sensitizing the CSC to denosumab and synergistically reducing the cancer initiating cell population and their capacity for self-renewal.
Breast CSCs, MDA-MB-436 [177] Dichloroacetic Acid (DCA) DCA and metformin when used in combination synergistically induced caspase dependent apoptosis in cancer cells. Metformin-associated oxidative stress-induced damage was amplified by DCA treatment associated pyruvate dehydrogenase kinase 1 inhibition thereby reducing metformin mediated lactate production.
MCF7, T47D, MCF10A [178] The combinatorial therapeutic strategy using DCA and metformin inhibited key glycolytic enzymes-hexokinase 2, lactate dehydrogenase A, and enolase 1. An activation of HIF1α abolished the effect of the combination therapy and reversed the inhibition on the expression of the glycolytic enzymes and reduced cell death.  DOX Nanoparticle co-encapsulated metformin and DOX achieved good tumor penetration, inhibited NF-κB activity, and decreased TNFα and IL6 expressions leading to the significant decrease in cancer cell proliferation. The nano-formulation of metformin and DOX showed a therapeutic effect in the treatment of lipopolysaccharide (LPS)-induced pulmonary metastasis model of murine 4T1 cells.
4T1, BALB/c mice inoculated with 4T1 cells [181] Nanoparticle co-encapsulated metformin and DOX (dual drug loaded) treatment showed increased toxicity and apoptotic cell death in DOX resistant MCF7 cells. The enhanced efficiency and cytotoxicity were attributed the to the intracellular accumulation of the drugs via enhanced cellular uptake and reduction in drug efflux leading to significant energy stress (reduced cellular ATP) and inhibition of multidrug resistance (MDR) mediating P-glycoprotein (P-gp).
MCF7, MCF7/DOX resistant cells [182] Metformin and DOX dual drug-loaded nanoparticles effectively reduced P-gp expression and activity, increased energy stress as evidenced by reduced intracellular ATP levels, and sensitized the cells to DOX induced apoptotic cell death. MCF7, MCF7/DOX resistant cells [183] Metformin treatment associated AMPK dependent anti-tumor effect was observed in addition to the inhibition of NF-κB and cyclin D1 gene expression. Combinatorial treatment was found to be more effective in decreasing tumor volume and improve overall rate of survival in the animals. Higher rates of apoptosis were observed in histopathological samples derived from animals to which the combination treatment was administered. Metformin treatment-associated reduction in the P-gp expression and elimination of Ki-67 positive cancer cells were observed in MCF7/ADR tumor xenografts.
The Ehrlich ascites carcinoma cells (derived from mouse breast adenocarcinoma cells) were implanted and allowed to multiply in the peritoneal cavity of Swiss albino mice. Solid Ehrlich carcinoma were derived by implanting EAC cells subcutaneously in Swiss albino mice. [184] The combinatorial treatment synergistically reversed DOX resistance both in vitro and in vivo. Metformin inhibited tumor growth. The cytotoxic effects of metformin were enhanced by increasing the levels of ROS while the levels of ATP levels depleted.

MCF7/ADR cells-DOX resistant cells, subcutaneously implanted MCF7/ADR cells in nu/nu mice [185]
Erlotinib The combination treatment synergistically induced apoptotic cell death and reduced the phosphorylation of EGFR, Akt, S6, and 4EBP1, prevented colony formation and inhibited mammosphere outgrowth. Everolimus Metformin-induced additive effects were observed when used as a co-treatment with everolimus and inhibited cell proliferation and colony formation ability. The additive effect of the combinatorial treatment was also related to the inhibition of mitochondrial respiration and mTOR growth signaling.
MCF7, MDA-MB-231, T47D [187] Inhibition of cell proliferation and tumor growth was observed both in cultures and mouse xenograft models treated with a combination of everolimus and metformin. Significant decrease in the levels of phosphorylated S6 ribosomal protein and 4E-BP1 was observed upon combination treatment.
HCC1428, MDA-MB-468, BT549, BALB/c mice inoculated with HCC1428 cells [188] Flavone Significant inhibition of cell viability, increased apoptosis, decrease in the expression of murine double minute X (MDMX), and activation of p53 via the PI3K/Akt pathway was observed in the combination drutreated cells. Apoptosis was mediated by decrease in Bcl2 and increase in the levels of Bax and caspase 3.
MCF10A, MCF7, MDA-MB-231 [189] Melatonin DMBA induced tumor incidence, tumor growth, and volume were reduced by the combination treatment. The apoptotic stimulation observed in the cancer cells was attributed to the activation of caspase 3.
7, 12-dimethylbenz[a]anthracene (DMBA) induced in vivo rat model of breast cancer [190] Paclitaxel (PTX) Co-delivery of PTX and metformin using a folate-modified amphiphilic and biodegradable biomaterial synergistically decreased cell proliferation and induced apoptosis through the toll-like receptor (TLR) signaling via the modulation of the TLR-MyD88-ERK pathway (responsible for tumor growth, progression, metastasis, and drug resistance).
4T1 cells, BALB/c mice inoculated with 4T1 cells [191] Silibinin Synergistic effect on growth inhibition of cancer cells was observed when silibinin and metformin were used in combination. Downregulation of hTERT and cyclin D1 was observed with the combinatorial therapeutic approach. T47D [192] Spautin-1 Deletion of the essential autophagy gene, Rb1cc1, suppressed tumorigenesis in BRCA1-deficient mice, while tumor growth and distribution of histological subtypes were not affected by loss of Rb1cc1. Co-treatment using spautin-1 (autophagy inhibitor) and metformin (mitochondrial complex-1 inhibitor) efficiently reduced the oxidative respiratory capacity, colony forming ability, and tumor growth.

Tamoxifen
The dosage of tamoxifen required for growth inhibition of cells was much lower when combined with metformin than when used as a monotherapy. The combination treatment inhibited cellular proliferation, DNA replication activity, colony formation, and activated apoptotic cell death in ER+ve breast cancer cells. The involvement of the Bax/Bcl2 apoptotic pathway and the AMPK/mTOR/p70S6 growth pathways were implicated in the beneficial effects of this combinatorial therapy approach.
MCF7, ZR-75-1 [194] Topotecan Metformin and topotecan dual drug carrier nanoparticles were found to be synergistically cytotoxic for the breast cancer cells, effectively promoting cell death via mitochondrial membrane depolarization and cell cycle arrest. MDA-MB-231, 4T1 [195] Vitamin D3 The combination treatment of metformin and vitamin D3 in synergistically inhibited cell proliferation and activated apoptosis in breast cancer cells. Mechanisms involving activation of AMPK, upregulation of Bax, cleavage of caspase 3, and inhibition of pBcl2, c-Myc, pIGF-IR, pmTOR, pP70S6K, and pS6 were implicated in anti-cancer activity of the combinatorial treatment.
MDA-MB-231 [196] Radiation A higher tumor response to radiation was observed in diabetic breast cancer patients who received metformin and partly yielded survival benefits. Meta-analysis/Clinical [197] 6. Challenges, Future Perspective, and Directions

Mixed Messages and Challenges
The number of epidemiological and meta-analysis studies that support the efficacy of metformin as an anti-cancer/anti-tumor agent have significantly increased ever since diabetologists reported the unusually low rate of cancers among diabetic patients treated with metformin. It was and is still being followed by attempts to provide well defined scientific evidences through lab-based and clinical studies to repurpose metformin and establish it as a drug that can be used in the treatment of wide variety of cancers, including breast cancer. While the identification of new and safer anti-cancer agents is always welcome in the medical field, metformin has garnered a special interest and significance in this regard since it is well tolerated and safe, off patent regulations, easy to synthesize, and can be made available cheaply and extensively in the global market.
In spite of all the studies and data that supports the use of metformin as an anti-cancer agent in breast cancer, certain studies have failed to show any significant association between the use of metformin and the stage of breast cancer, characteristics of the tumor at diagnosis, and survival of patients [162,163]. Patients receiving concurrent metformin and radiation experienced increased locoregional toxicity, higher frequency of treatment breaks, and desquamation/dermatitis despite the fact that several preclinical studies have demonstrated the metformin treatment-associated sensitization to radiation therapy [198][199][200]. Reports have also identified time-related biases in the observational studies that systematically tend to exaggerate the reported anti-tumor effects of metformin [201,202].
Several in vitro and in vivo studies have come up with significant data supporting the use of metformin as an anti-cancer agent in breast cancer, as a monotherapy, and, with even better efficacy when used in combination with other routinely used chemotherapeutic drugs/radiation therapy and or other naturally occurring compounds with known anti-cancer potential. Surprisingly, although several clinical trials (44 breast cancer and metformin treatment-related trials, as discussed above) have been undertaken to provide clinical evidence-based support to the lab-based studies, out of the 14 that have been completed, only 6 have reported the data and published the results. The outcomes of ongoing clinical trials should, in future, provide more concrete answers with regards to the use and efficacy of metformin, in the prevention and treatment of breast cancer in diabetic patients, as well as its safety when used in non-diabetic breast cancer subjects [149].
There are several challenges that 'metformin' should face and overcome before it can be universally accepted for its anti-cancer potential, in a manner it is currently accepted and used as an anti-diabetic drug. Dosage of metformin administration as a monotherapy or in combination with other drugs or therapeutic modalities is critical to achieve therapeutic efficiency in the treatment with minimal side effects. When used orally in the treatment of diabetes, the anti-hyperglycemic effects of metformin have been reported at plasma concentrations ranging from around from 10-100 µM [68,137,203]. Interestingly, epidemiological, observational, and meta-analysis reports suggest that oral administration of metformin (normal initial oral dose for; 'immediate release' is 500 mg twice/day or 850 mg once/day, with 500 mg increments weekly as tolerated, maximum dose of 2550 mg/day; 'extended release' is 500 to 1000 mg once/day, with 500 mg increments weekly as tolerated, maximum dose of 2000 mg/day) in type 2 diabetic patients significantly reduced the risk of many different cancers when compared to diabetic patients on other anti-hyperglycemic treatment regimens [35,37,204,205]. On the other hand, some studies have exercised caution while interpreting observational studies and systematic reviews, suggesting that factors such as the time-related bias were not considered leading to the inference that metformin treatment led to significant (ranging from 20% to 94%) reduction in the risk/incidence of cancer [201,206]. Metformin use showed no effect on the incidence of cancers in studies that avoided these biases [201,206]. In the majority of in vitro/cancer cell based experiments, however, metformin treatment-related inhibitory effect on tumor cell proliferation, activation of apoptotic cell death, and inhibition of tumor progression was only observed at >2-5 mM concentrations of metformin [68,137,138,203,207]. Further studies are warranted to provide a clear insight into the possible dose-dependent effect of metformin on cancer prevention/risk reduction in 'diabetic' patients and cancer treatment in 'diabetic-cancer' patients. It may be, however, necessary to manage diabetes among cancer patients since the efficacy of metformin as an anti-cancer agent is largely dependent on the concentration of glucose in the tumor microenvironment [5,149]. Pharmacological deprivation of glucose/inhibition of energy deriving glucose metabolism in cancers when combined with metformin treatment should be more beneficial in cancer treatment, thereby eliciting an improved patient response to the therapeutic intervention [5,149].
Additionally, it is still unclear how these massive concentrations can be achieved in a clinical setting when used in the treatment of cancer and whether there would be dosage dependent toxicity related adverse effects that may outweigh the beneficial effect of metformin as an anticancer agent. Although few studies on the metformin transporters, OCTs, PMAT, and MATEs, provide insight into the possible intracellular accumulation of metformin in cancer cells, more studies are warranted in this area to ascertain the specificity to cancer cells and whether the normal non-cancerous cells will be spared from these effects of metformin depending on the absence or presence or the ratio of the metformin intake or extrusion transporters. Additionally, the heterogeneity of the transporter expression in different cancers must be carefully studied and addressed in this regard to arrive at a safe and effective dosage of metformin when used in the treatment of cancer. Studies using metformin in combination with other chemotherapeutic drugs or natural compounds with anti-cancer potential may hold the key to identifying synergistic effects of the combination therapy, thereby effectively decreasing the dosage required with better therapeutic efficiency when compared to usage as a monotherapy in the treatment of cancer.
It is true that several cell-based and in vivo studies have shown the several different molecular mechanisms that support the anti-tumor potential of metformin monotherapy. Such studies are indeed very important to elucidate molecular mechanisms and identify therapeutic targets of metformin. In clinical trials (Table 2), however, metformin is rarely studied as the sole therapeutic agent, but is used as co-treatment with other routinely used chemotherapeutic drugs [208]. This is indicative of the fact that it is unlikely, although not impossible, that metformin will gain importance as a monotherapy in cancer treatment. Nonetheless, metformin surely has the potential as an anti-cancer drug, which can be accepted and be used as neoadjuvant/adjuvant therapy or as co-treatment in combination with routine anti-cancer treatment regimens [208].
A vast number of studies suggest that metformin co-treatment can overcome resistance to chemotherapeutic drugs/radiation and can sensitize the cancer cells to the ant-cancer agents/radiation. However, like any other chemotherapeutic drug/treatment modality, long-term metformin administration can also be challenged with the evolution of treatment-related resistance. In response to long-term metformin treatment, MCF7 breast cancer cells reportedly developed cross-resistance to metformin and tamoxifen and was dependent on the constitutive activation of Akt/Snail1/E-cadherin signaling [209]. As discussed earlier in the article, combinatorial therapeutic approaches are key to avoiding treatment-related drug resistance.

Metformin and Breast Cancer Biomarkers
In cancers, identification of suitable diagnostic and prognostic biomarkers is critical to screening for, early detection of, development of therapeutic strategy for, and evaluating the responsiveness to the therapeutic intervention of the disease. ER and PR hormone receptors, BRCA1 and BRCA2 gene mutations, HER2/neu gene amplification or protein overexpression, plasminogen activator inhibitor 1 (PAI-1), urokinase-type plasminogen activator (uPA), 70-gene signature (Mammaprint ® ), and 21-gene signature (Oncotype DX ® ) are being currently used as markers for breast cancers [5,210,211]. More recently, the cancer testis antigens (CTAs) that are in physiology normally expressed in the testis and placenta are expressed in different forms of cancers and have been implicated in tumorigenesis, thus gaining importance as diagnostic molecular signatures with prognostic value in cancers including breast cancers [5,212,213]. CTAs are involved in cellular proliferation, migration/motility, colonization, and cell division in normal germ cells, while in cancer cells, CTAs have been implicated in sustenance of growth, maintaining nutrient and oxygen supply through the activation of angiogenesis, evading apoptotic cell death, increase in tissue invasiveness and metastasis, and development of resistance to anti-cancer drugs [214]. In breast cancer, CTAs, such as MAGE-1, AKAP4, NY-BR-1, CTAG1, BAGE1, MAGE-A10, SP17, NY-ESO-1, and MAGE-A, were found to play important roles in activating cell proliferation, inhibition of apoptosis, and promoting tissue invasion and metastasis [5,[215][216][217][218][219][220][221][222]. The CTA, AKAP4, in particular seems to be an interesting candidate for a serum-based diagnostic test for the early detection and diagnosis of breast cancer [221]. SP17, MAGE-A, NY-ESO-1, and MAGE-A10 on the other hand were more specifically associated with TNBCs [216,218,219]. There is but little evidence that could link the occurrence of diabetes to the expression of the CTAs, which can then be possibly used as diagnostic markers to link diabetes and the incidence of breast cancer or even a specific kind of breast cancer. Moreover, it is also necessary to evaluate if there is any association between the anti-hyperglycemic treatment used in diabetes, such as insulin or metformin, and the expression or levels of specific CTAs in breast tissue, which could ultimately also serve as prognostic markers to study the response of the cancer to metformin intervention.

Efficacy in Non-Diabetic Patients and Non-Diabetic Cancer Patients
As discussed, several observational studies, systematic reviews, and meta-analysis studies report the ability and efficacy of metformin in risk reduction of many types of cancer, including breast cancer, in type 2 diabetic patients on a metformin treatment regimen. Moreover, women using metformin in combination with other anti-hyperglycemic drugs reportedly showed a lower risk of breast cancer occurrence, although not significant, when compared to women using metformin alone [223]. While metformin notably inhibits cell growth and proliferation-related pathways, these effects may be in part related to the ability of metformin to reduce insulin resistance, insulin levels, and circulating levels of glucose [65]. However, it remains largely unknown if non-diabetic women with breast cancer, or non-diabetic individuals with other forms of cancer, would benefit from taking metformin [149].
Medical science has advanced in leaps and bounds in the area of cancer treatment, with research teams constantly searching for and introducing newer and more efficient drugs/treatment modalities for the treatment of cancer. Nonetheless, in a world where cancer is one of the most feared and dreaded diseases, it will depend on the results of the currently active clinical trials to determine and recommend whether it is feasible and beneficial to administer metformin (1) as a cancer 'preventive' drug in non-diabetics to reduce the risk of cancers or (2) as an intervention, either alone or in combination with other anti-cancer drugs, for the treatment of cancer 'non-diabetic' cancer patients [149,224]. While using metformin as a preventive drug may decrease the occurrence of cancers in non-diabetic individuals, in non-diabetic cancer patients, metformin should increase the efficacy of anti-cancer drugs and show improved prognosis.

Future Directions-Metformin and a Proposed Path Towards Precision Medicine
Metformin has gained global attention for its potential as an anti-cancer drug and is currently being tested in in vitro cell-based assays, in vivo animal experiments, and in clinics at many cancer research facilities around the world. Although the end results are 'connected' in terms of molecular mechanism of action, inhibition of cell proliferation, activation of cancer cell death, and tumor suppression, there seems to be 'disconnect' when it finally comes from the bench to bedside application. It is true that targeting a unified mechanism makes it easier to treat a disease. However, the heterogeneity and variation in diseases such as cancer calls for more 'personalized' and 'precise' case-by-case approach in treatment modalities for a better outcome.
Not ignoring the fact that basic biomedical research will always remain extremely important, research resources must be channeled into this area of precision medicine that will be tailored to fit a patient based on the risk of disease and predicted response to the therapeutic intervention. When considering metformin for its potential cancer preventive or cancer-treatment effect, adhering to an approach of precision medicine and adopting a workflow (Figure 3) that will integrate information, knowledge, skills, and research data at all levels of research will certainly be beneficial to ultimately determine if the answer to cancer can indeed be found in a flower. The publication of this article was funded by the Qatar National Library.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the article.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the article.