1. Metformin
Metformin hydrochloride is the first-line agent for T2D, a disease which currently affects over 400 million people around the globe
[1]. Though currently the efficacy of metformin for treating T2D is seldomly questioned, its debut to the medical field initially spurred a widespread skepticism, mostly from the failed attempts of phenformin, a similar compound, to achieve an efficacy and safety profile adequate for clinical use
[2]. Currently, metformin is being studied for its action in T2D, neurodegenerative diseases, infectious diseases, and neoplastic diseases. The pleiotropic effects of metformin are vast, and though the current literature describes significant achievements, the full extent of metformin´s mechanisms in the context of various diseases are far from established
[3][4].
Metformin is a positively charged compound, therefore it requires the expression of specific proteins for its transport across biological membranes. Organic cation transporters (OCTs) have been shown to aid in the intracellular transport of metformin from plasma, particularly OCT-1, which is highly expressed by cells in organs including the liver, kidney, and intestine, and OCT-3 and MATE1 have also been shown to play a role in metformin transport. OCT expression reduction could decrease the antitumor effect of metformin
[2][5][6][7]. The distribution of metformin from pharmacokinetic studies has been consistent with the tissue-specific expression of the aforementioned transporters
[2]. In terms of pharmacokinetics, metformin is currently administered for T2D at doses ranging from 1000–2000 mg daily, and both an immediate release and a prolonged release formulation are available, with a similar oral bioavailability. Such doses lead to plasmatic metformin concentrations of approximately 10–40 µM, though as mentioned earlier, specific tissues will have higher concentrations from preferred uptake, such as the liver
[8][9][10]. It is important to highlight, particularly regarding the effects of this research, that in non-diabetic patients, treatment with 1000 mg of metformin will achieve a plasmatic concentration of 25 µM within 3 h of administration
[2].
The glucose-lowering effects of metformin were established in several basic studies, having a direct effect on the endogenous production of glucose by the liver, and this effect was mediated through a tight regulation of the gluconeogenic pathway, while glycogenolysis remained mostly undisturbed
[2][11][12][13].
Most basic studies regarding the mechanism of action of this agent have reached the conclusion that metformin inhibits the mitochondrial electron transport chain (ETC), particularly by disrupting complex I, which is the site of entry for electrons moved through reduced nicotine adenine dinucleotide (NADH),
[14][15][16]. The inhibition of the ETC in turn would decrease adenosine triphosphate (ATP) production, NADH oxidation, and oxygen consumption, and this would trigger a cellular response to attempt to adapt to these new bioenergetic conditions, increasing expression of glycolytic enzymes and hampering glucagon signaling. As ATP levels decrease, AMP levels rise and activate adenylate cyclase. Further, this energy shortage would also activate 5′-AMP-activated protein kinase (AMPK), an energy sensor that stimulates catabolic pathways for energy generation (
Figure 1)
[15]. However, it is important to highlight that, recently, this mechanism of action has been called into question, and although all the reviews currently identified build on this mode to explain the diverse effects of metformin treatment in various illnesses, it must be stated that the observations that led to establishing complex I inhibition by metformin were achieved using metformin concentrations which surpassed the pharmacological bioavailability achieved using the currently approved posology by dozens or even hundreds of times (millimolar concentrations).
Figure 1. Metformin effects on non-transformed cells versus lung cancer cells. Metformin treatment has different effects on LC cells than non-transformed cells. In both cell types, metformin exerts its effect through a dysfunction of complex I of the electron transport chain, however, in non-transformed cells, metformin promotes glucose incorporation through increased expression of GLUT1 (green membrane proteins), hexokinases (orange circles), and decreasing the gluconeogenesis process. The ATP generation process is also increased through activation of AMPK and with a sustained OXPHOS. In LC cells, metformin stimulates energy generation through over-activation of OXPHOS processes associated with an increasing of AMPK pathway activity, mTOR inhibition, and a decrease in all glycolytic proteins and processes that are associated with this metabolic pathway. (Created with BioRender.com).
Interestingly, this premise that metformin inhibits the hepatic ETC complex I is the foremost used explanation behind the potential anticancer activity of metformin treatment
[17]. AMPK is a highly relevant serine/threonine protein with kinase activity which is located in the cytoplasm; this kinase has three subunits (the catalytic α-subunit, the scaffolding β-subunit, and the regulatory γ-subunit). AMPK activation occurs mainly through an increase in AMP/ADP ratio and by small molecules which mimic AMP. The AMPK complex undergoes a conformational change that improves Thr-172 phosphorylation, which in turn leads to AMPK activation, increasing catabolic pathways, and decreasing anabolic pathways
[18][19].
One of the kinases which can phosphorylate and activate AMPK is liver kinase B1 (LKB1), and a study on LKB1 knockout mice showed the animals present with hyperglycemia and inactivation of AMPK and are resistant to the therapeutic effect of metformin. Some studies have shown that AMPK activation is achieved through a signaling loop by AMP binding. AMPK activation allows interaction with AXIN–LKB1, generating a ternary complex AXIN–LKB1–AMPK, which facilitates the phosphorylation of AMPK by LKB1
[18][20].
On the other hand, AMPK activation is also dependent on intracellular calcium concentration, and high calcium concentrations have been shown to promote calcium/calmodulin-dependent protein kinase kinase-β (CaMKKβ) activation, leading to AMPK phosphorylation
[21][22]. Accordingly, metformin indirectly activates AMPK through the secondary activation of LKB1. LKB1 phosphorylates the catalytic α domain of AMPK and its activity is stabilized by the formation of a heterotrimeric complex with pseudo-kinase STe20-related adaptor (STRAD) and the mouse 25 protein (MO25). Once the complex is formed, LKB1 activates cell signaling pathways of cell growth regulation and catabolism
[23][24]. On the other hand, LKB1 orchestrates a cellular response to energetic status and regulates the balance between catabolic and anabolic processes
[25]. In this regard, AMPK activation by metformin in these studies leads to a transcriptional downregulation of gluconeogenic genes as well as the phosphorylation of acetyl-CoA carboxylase (ACC) which reduces lipogenesis and promotes hepatic mitochondrial beta-oxidation
[2]. In fact, some researchers have claimed that AMPK and LKB1 activation are key components for metformin to inhibit glucose production in hepatocytes and stimulate glucose uptake in skeletal muscles
[26]. Further, studies have shown that metformin at clinically relevant plasma concentrations does not appreciably affect energy charge or AMP concentrations, in fact, the therapeutic effects of metformin have been observed even in the absence of AMPK expression
[27]. Therefore, it is unlikely based on current literature that AMPK activation is fundamentally required for metformin action in terms of the metabolic effects of the drug
[2].
Recently, an elegant series of experiments performed by Madiraju et al. have shown that the most likely primary molecular target of metformin is glycerol 3-phosphate dehydrogenase (GPD2), an enzyme necessary for glycerol to enter the gluconeogenic pathway
[27]. As a three-carbon skeleton, glycerol may enter the gluconeogenic pathway depending on the cellular redox state (NADH: NAD + ratio), and metformin inhibits GPD2 activity and alters the cytosolic redox balance, suppressing the entry of redox-dependent substrates to the gluconeogenic pathway (in addition to glycerol, lactate also depends on redox state for entry to gluconeogenesis)
[27]. This substrate-specific mechanism of action explains in part the low rate of hypoglycemia observed with metformin treatment, as well as the abrogation of metformin action by infusion of methylene blue (which normalizes cytosolic redox state). Moreover, this would imply that metformin may have a disproportionate benefit for individuals with dysregulated white adipose tissue lipolysis
[2], and one condition in which lipolysis is dysregulated is obesity, in which an enhanced baseline lipolysis is observed, likely due to impaired sensitivity of adipocytes to hormonal signaling
[28]. This novel observation would imply that the effects of metformin are different in lean and obese subjects, an observation established several years ago in the seminal studies by DeFronzo regarding the effects of metformin for muscular insulin-stimulated glucose uptake. In this study, metformin increased whole body insulin-stimulated glucose uptake, but the effect was exclusively observed in patients with T2D and obesity, highlighting the paradoxical effects of metformin in lean vs. obese individuals
[28][29].
2. Molecular Effects of Metformin in Lung Cancer
As previously mentioned, several studies have identified an increased risk of lung cancer for patients living with diabetes, and although results have been inconsistent, some studies have further shown that patients with lung cancer and diabetes have overall a worse prognosis
[30].
It is important to highlight that studies have shown inconsistent results on LC prognosis among patients with this comorbidity
[30][31][32], and as such definitive conclusions have not been reached regarding this association. These contradictory results might be related to a heterogeneous population, patients with different stages or with different glycemic control, and different antidiabetic treatments.
Interestingly, among patients with T2D and cancer, studies have more consistently shown that treatment with metformin, but not with other hypoglycemic agents, has beneficial effects on survival; results further show that metformin administration is associated with a reduction in cancer risk and cancer-associated mortality
[33][34]. An observational study by research group reported that in patients with NSCLC and T2D, treatment with metformin is associated with an improvement in overall survival compared with patients not treated with metformin
[35]. In another study, patients with T2D and SCLC treated with metformin also showed an improvement in both overall survival and progression-free survival (PFS) compared with patients that were not treated with metformin
[36]. These results suggest that metformin could improve the prognosis of patients with T2D and lung neoplasms.
In non-transformed cells, metformin reportedly works through acceleration of the glucose assimilation–consumption processes and ATP formation (a mechanism mostly thought to be mediated through the AMPK signaling pathway). In accordance with this idea, it has been suggested that the antitumor effect of metformin could also exploit the AMPK axis (
Figure 1). However, the antiproliferative effects of metformin are also observed in LKB1-null melanoma cells, and therefore it must be acknowledged that the breadth of scope of the diverse anticancer pathways that could be affected by metformin treatment is still mostly unknown
[37]. Further, metformin has been shown to induce cell cycle arrest and apoptosis by inhibiting mTOR activity, a process which is independent of AMPK activation
[37]. Therefore, the antineoplastic effect of metformin could be mediated by AMPK-dependent mechanisms and AMPK-independent mechanisms.
2.1. AMPK-Dependent Mechanisms
Tumor cells have metabolic alterations that confers upon them a decreased ability to respond to an energy-deprived status, creating a metabolic vulnerability that could be used as a therapeutic approach. Several LC models have described an increase in energy generation through glycolytic processes and a decrease in ATP generation through oxidative phosphorylation (OXPHOS). Thus, the metabolic Warburg effect (WE) phenotype characteristic of LC has been associated with deregulation of signaling pathways such as the LKB1–AMPK axis.
Initially, LKB1 deficiencies were detected in Peutz–Jeghers syndrome, a disease that increases the risk of suffering pleiotropic types of cancer
[38][39]. Then, it was demonstrated that several lung tumors display deletions of chromosome 19, where the LKB1/STK11 gene is located
[40]. Furthermore, thirty percent of patients with advanced somatic KRAS-mutant NSCLC show LKB1 deletions that lead to a partial or complete inactivation of protein function
[41]. KRAS and LKB1 co-mutation confers worse outcomes in murine models and is associated with resistance to selumetinib and docetaxel
[42].
At a molecular level, KRAS and LKB1 co-mutation increases the production of reactive oxygen species (ROS) and decreases ATP generation, NADPH/NADP + ratio, and glutathione levels. This process is strongly associated with activation of the KEAP1/NRF2 pathway which increases cell survival
[43]. Moreover, cancer cells deficient in LKB1 show increased autophagy activity, which provides fuel for the formation of mitochondrial energy through autophagosome formation
[44].
Towards a different molecular pathway, AMPK activation upon metformin exposure contributes to the upregulation of peroxisome proliferator-activated receptor-gamma coactivator 1-alpha (PGC-1α). PGC-1α is a transcriptional coactivator responsible for mitochondrial biogenesis; it has been reported that low levels of PGC-1α are associated with worse OS
[45][46]. Metformin increases PGC-1α and prevents gluconeogenesis activation
[47]; altogether, these data indicate that loss of LKB1 and high expression of AMPK modify lung cell metabolism to generate energy for uncontrolled proliferation.
Table 1 summarizes how metformin treatment modifies AMPK expression and activation in several LC models.
Table 1. Overview of studies regarding use of metformin in cancer cell lines.
2.2. AMPK-Independent Mechanisms
Antitumor effects of metformin could also be mediated by AMPK-independent mechanisms. Several reports indicate that AMPK is not required for the control of glycemia. In a previous study, treatment with metformin inhibited the proliferation of LKB1- and AMPK-null cancer cells, highlighting a potential AMPK-independent mechanism of action for antiproliferative effects. An additional study corroborated these findings, reporting that metformin inhibited proliferation of LKB1-positive H1299 cells and LKB1-null H460 cells, suggesting that LKB1 is not necessary for this activity. Three reports indicated that the effect of metformin on cancer cells does not require activation of AMPK: (1) in prostate cancer cells, downregulation of AMPK did not affect metformin action and mTOR inhibition
[21]; (2) the activation of AMPK was not required for the antimelanoma action of metformin and the use of an AMPK inhibitor failed to restore viability in metformin-treated cells
[54]; and (3) inhibition of glucose production following treatment with metformin occurred in both AMPK- and LKB1-deficient hepatocytes
[22]. These data indicate that metformin could act by an alternative pathway in specific cells and underscore the complexity of mechanisms activated by this compound.
3. Metformin in Lung Cancer Therapy
The effects of metformin on LC as a single agent have been studied in vitro and in animal models. Metformin treatment in LC cells inhibits proliferation, induces apoptosis, and decreases colony formation. However, most of the currently published studies used metformin at millimolar concentrations, while concentrations of metformin in blood samples of patients treated with approved schemes are in the micromolar range
[55]. In vivo studies using metformin in drinking water (5–250 mg/kg) demonstrated that treatment inhibits lung tumorigeneses and metastases through inhibiting both mTOR activation and phosphorylation of multiple tyrosine kinase receptors (EGFR, IGFR, and VEGFR) by AMPK-dependent and -independent mechanisms
[55][56]. These data provide insight into the potential of metformin to augment the efficacy of existing lung cancer therapeutics.
3.1. Metformin as an Adjuvant in Lung Cancer Therapy
A combination of metformin with other therapies not only regulates the LKB1–AMPK axis, it also stimulates and regulates several cellular signaling pathways depending on the type of combination treatment. For example, metformin can inhibit hormone pathways when it is combined with hormonal treatment, inducing cell cycle arrest and apoptosis. Studies of metformin’s combination with antimetabolic drugs such as 5-fluorouracil have shown inhibition of tumor proliferation, suppression of hypoxia-inducible factor 1-α (HIF-1α), and downregulation of multidrug resistance-associated protein 1 (MRP1). In addition, when used in combination with chemotherapeutic drugs, metformin inhibits lipogenesis and cholesterol synthesis through ERCC1 downregulation, promoting tumor cell death
[45][57][58][59].
3.1.1. Metformin plus Platinum-Based Chemotherapy and Radiotherapy
NSCLC patients with either absence of a targeted oncogenic driver mutation or high programmed death-ligand 1 (PDL-1) expression are frequently treated with platinum-based doublet chemotherapy and bevacizumab, particularly when access to immunotherapy is lacking
[60]. Median overall survival in this patient subgroup ranges from 10 to 12 months, demonstrating the need for new therapeutic opportunities.
The potential effect of combining metformin with several chemotherapeutic agents has been tested in NSCLC cell lines. Chemoresistance to cisplatin is related to ROS production, IL-6 secretion, and signal transducer and activator of transcription (STAT)-3 phosphorylation. Metformin increases the antitumor effectiveness of platinum-based chemotherapy and inhibits cisplatin-induced ROS and IL-6 secretion through modulation of the STAT3 pathway by an LKB1–AMPK-independent mechanism
[61][62][63]; these studies portray the potential of combining metformin with standard first-line platinum-based chemotherapy to improve efficacy.
Several clinical trials have examined the effect of platinum-based chemotherapy combined with metformin. Initial clinical studies showed an association between metformin and decreased recurrence or progression in patients with LC and diabetes. In a retrospective study, 143 patients with LC with a pre-existing diabetes diagnosis were assessed for outcomes; the subgroup of patients treated with metformin displayed a tendency toward a better disease control rate. In addition, patients treated with insulin had a worse tumor response
[64]. In the same study, metformin plus chemotherapy significantly increased PFS and overall survival (OS) (20.0 months vs. 13.1 months vs. 13.0 months, respectively;
p = 0.007) when compared with insulin and drugs other than metformin and insulin.
An open-label phase II study enrolled patients without T2D with chemotherapy-naïve advanced or metastatic non-squamous NSCLC to receive carboplatin, paclitaxel, and bevacizumab with or without concurrent metformin (NCT01578551). The primary analysis showed that the addition of metformin improved 15% of PFS in the first year, compared with patients treated with carboplatin, paclitaxel, and bevacizumab as monotherapy. The research was stopped early due to changes in practice patterns for non-squamous treatment. However, it reported an increase in OS in patients that received metformin (15.9 vs. 13.9). The safety profile for patients undergoing the combined therapy did not highlight adverse effects attributed to metformin. On the other hand, one retrospective study reported that the combination of metformin with platinum-based chemotherapy was not associated with any survival benefit in patients with NSCLC
[65][66].
Recently, a hypothesis emerged suggesting that nutrient deprivation could enhance the effects of metformin during chemotherapy. Based on this idea, the phase II FAME trial was designed. Patients enrolled in this trial will receive platinum-based chemotherapy and pemetrexed with metformin alone or metformin plus a fasting-mimicking diet. It is estimated that the two treatments shall improve median progression-free survival from 7.6 months to 12 months
[25].
In addition to its role in combination with chemotherapy, pre-clinical studies have shown that metformin enhances radiosensitivity through activation of ataxia–telangiectasia mutated (ATM) gene product and AMPK
[67]. ATM is an important signal transduction cascade for repairing DNA damage, and establishes the degree of radiosensitivity or radioresistance. This improvement of response to radiotherapy mediated by metformin has been suggested to occur through a downregulation of the hyperactive PI3K–AKT–mTOR pathway
[6].
Some retrospective cohort studies have tested the hypothesis that metformin enhances radiosensitivity in NSCLC. A study reported that patients with advanced NSCLC that received metformin plus radiotherapy (50 Gy administered in a fraction of 1.5–2.75 Gy once or twice daily), along with cisplatin, have a median overall survival of 62 months; meanwhile patients in the control group (radiotherapy plus cisplatin) only achieved a median OS of 49 months. Furthermore, distant metastasis-free survival and progression-free survival were better in the patients who received metformin
[68][69].
The randomized phase II trial ALMERA compared standard radiotherapy plus concurrent chemotherapy with or without metformin in unresectable, locally advanced NSCLC patients without diabetes (NCT02115464). Treatment consisted of scaled doses of metformin orally, in the first week receiving 500 mg twice per day, in the second week 1500 mg, and in the third week 2000 mg and platinum-based chemotherapy with or without consolidation with standard radiotherapy (60–63 Gy) for six weeks. Though the research was well designed to assess the benefit of metformin in this indication, it was stopped early due to slow accrual. Among the patients who were randomized from 2014–2019 (
n = 54), results showed that addition of metformin to chemoradiotherapy was associated with inferior efficacy and higher toxicity
[70]. Another study evaluated the use of metformin and/or insulin in NSCLC patients with diabetes mellitus undergoing radiochemotherapy; the study enrolled 70 patients, and although recruitment status is complete, results have not been published thus far (NCT02109549).
Initial reporting of NRG-LU001 (NCT02186841), a randomized phase II trial of concurrent chemoradiotherapy plus metformin in advanced NSCLC, showed no significant differences in rates or grade of toxicity between groups. Metformin did not improve PFS or OS and did not alter distant metastasis. Altogether, the data indicate that the benefit of metformin in terms of radiation therapy may be circumscribed to a specific disease stage and biomarker profile (i.e., LKB1 status), warranting further research.
3.1.2. Metformin plus TKIs
Several studies have evaluated the effect of adding metformin in regard to efficacy of EGFR tyrosine-kinase inhibitors (TKIs) and therapeutic resistance in LC. It has been reported that metformin can synergize with gefitinib, inhibiting cell growth, and reducing AKT/PI3K/mTOR pathway activity in LKB1 wild-type NSCLC cell lines
[71]. Metformin–TKI treatment was tested in transformed TKI-resistant LC cell lines with epithelial–mesenchymal transition (EMT) patterns, and the treatment showed sensitization of TKI-resistant cell lines promoting EMT regression, increasing adhesion cadherins and inhibiting IL-6 signaling
[72]. Recently, researchers' group reported a synergic effect between metformin and afatinib, and this effect was associated with a reduction in the EGFR pathway, glycolytic, and EMT markers
[73]. In contrast, another study described an increase in EGFR protein expression with metformin monotherapy, but when combined with erlotinib, metformin showed synergism and growth inhibition of EGFR wild-type LC cells
[74]. In TKI-resistant cell lines with mutant EGFR, a combination of metformin with gefitinib inhibited p-IGFR and p-AKT, and showed a synergic effect in apoptosis induction
[57]. It has also been observed that EGFR inhibition reactivates OXPHOS, and reverts WE in LC cells.
3.1.3. Metformin plus Immune Checkpoint Inhibitors
Immunotherapy has demonstrated a significant benefit in NSCLC patients
[75][76]. Studies in mouse models showed that the addition of metformin to nivolumab increased CD8 + tumor-infiltrating lymphocytes and decreased the production of interleukin 2 (IL-2), tumor necrosis factor (TNF), and interferon-gamma (IFN-γ) and, altogether, this protects lymphocytes from apoptosis and exhaustion. Metformin further prevents the apoptosis of CD8 + TIL in TME regardless of PD-1 or Tim-3 expression and promotes tumor cell rejection. Additionally, metformin and phenformin inhibit myeloid-derived suppressor cells (MDSCs) and, as a result, increase the antitumor activity of PD-1 blockers
[77]. It is also worth mentioning a recent study that identified that obesity could shape the metabolic profile of cells (both cancer and immune cells) in the tumor microenvironment and suppress antitumor immunity. Results from this research showed that cancer cells from diverse tumors may readily adapt in obese subjects to a lipid-based metabolic phenotype, essentially “starving” lymphocytes through a strategy to improve their uptake of lipids and depriving immune cells in this nutrient competition
[78]. As such, it remains to be evaluated whether metformin could be alleviating this adaptation through a differential nutrient profile which could benefit immune cells. Future studies in this regard might consider evaluating the role of PHD3, which is downregulated in tumor cells of obese subjects and is a plausible candidate for achieving this preferential metabolic profile by the neoplasm
[79].
3.1.4 Hypotheses on Metformin, Obesity, and Lung Cancer
The role of metformin in treatment of lung cancer has been difficult to establish thus far. Interestingly, metformin´s salutary effect on this particular neoplasm may play out similarly to its salutary effects in individuals with or without T2D and those with or without obesity. In this regard, early studies pertaining to the use of metformin to stimulate muscular glucose uptake showed metformin increased whole-body insulin-stimulated glucose uptake in patients diagnosed with T2D who presented with obesity [80]. Moreover, this effect could not be exclusively attributed to skeletal muscle uptake, and one possibility involves glucose uptake by adipocytes, which also have insulin-dependent glucose transporters (GLUT4). Clearly, in obese individuals, higher fat mass could in turn aid in glucose clearance. The studies overall indicate that metformin´s improvement of glucose uptake is an indirect consequence of improved glucose control from decreased hepatic glucose production, however, when metformin is used for treatment of non-diabetic individuals with lung cancer many such routes will not play out similarly, since such subjects already have an adequate glycemic control. However, it may be the case that metformin could change nutrient availability, speeding up glucose clearance by peripheral muscle and adipose tissue, thus leaving the tumor cells to achieve energy requirements from other nutrient profiles. Previous studies have shown that in non-diabetic individuals, metformin treatment using a hyperinsulinemic–euglycemic clamp produced increased glucose rate of disappearance and increased glucagon levels [81]. As a result of this profile, non-diabetic patients who are obese would initiate a signaling cascade to mobilize fatty acids stored in adipose tissue cells, through glucagon activation of hormone-sensitive lipase. Many tumors have been reported to readily adapt to lipid-based metabolism, which could predominate in individuals with high availability of such stored nutrients. In a recent study by Rimgel et al. [82], the researchers showed through an elegant single-cell sequencing strategy how tumor cells in obese subjects swiftly change metabolic patterns to improve fat uptake and result in tumor growth. This appears to be at least in part mediated through a decreased expression of PHD3, a prolyl hydroxylase which is relevant in regulating fatty acid oxidation; decreasing PHD3 expression by tumor cells can relieve the suppression of fatty acid transport to the mitochondria, resulting in increased beta-oxidation and ATP production.
However, if a specific tumor lacks this adaptation mechanism the outcome may be deleterious when faced with a higher availability of lipid-based nutrients for which tumor cells compete with other proliferative populations, such as CD8 + lymphocytes. Though PHD3 has not been extensively studied in lung adenocarcinomas, one recent report shows that in samples from surgically resected non-small cell lung cancer, PHD1 and PHD2 mRNA levels are decreased compared with normal tissue, but not PHD3 levels [83]. Whether PHD3 levels in tumors from lean and obese subjects are different remains to be explored, but would be plausible given the transcriptional regulation of this gene by axes such as glucagon–cAMP–PKA signaling in hepatocytes [84]. This could potentially explain the differences seen in the outcomes achieved from metformin treatment of obese vs. lean individuals and could be further explored as a biomarker to predict who could most likely reap benefit from this intervention.
4. Conclusions and Future Directions
Metformin is an oral antidiabetic medication that has been reported to regulate a key sensor of the energetic status of the cells and participate in the inhibition of cellular proliferation by impeding mitosis. Additionally, metformin modulates enzymes that regulate key metabolic pathways and lipid metabolism, leading to the downregulation of cell proliferation, migration, and protein synthesis. Among solid tumors, oxidative phenotype regression through metformin treatment has been broadly associated with a decrease in tumor growth. Metformin may exhibit anticancer properties through the regulation of growth factors such as EGFR and IGF. Metformin has also been shown to suppress epithelial–mesenchymal transition (EMT), impeding the spread of tumor cells. Clinically, the combination of metformin with tyrosine TKIs has been shown to improve the therapeutic response rate as well as OS in specific subgroups of NSCLC patients, particularly those with a high BMI. In addition, treatment with metformin can overcome therapeutic chemo- and TKI-resistance. However, the addition of metformin to conventional cancer therapy is still controversial, indicating that the potential benefit of this combination should be explored according to the novel data emerging regarding biomarkers and patient subgroups.
In this sense, it is necessary to find a potential biomarker to select patients that could benefit from metformin treatment. Loss of expression of LKB1 and AMPK have emerged as a response biomarker in NSCLC patients. Patients with loss of LKB1 with KRAS or EGFR mutations denote an aggressive tumor phenotype, that might be due to the loss of LKB1 as a metabolic sensor; these patients might benefit from biguanide treatment. In addition, AMPK is a potential metformin target, this receptor is lost in NSCLC, and recent reports show that activation of AMPK by metformin decreases PDL-1 levels, which in turn increases cytotoxic T cell activity against cancer cells. Lastly, the role of PHD3 which has emerged as an important regulator of the metabolic switch to lipids by specific tumors must be explored in this clinical context.
The question of which cellular processes and therapies can be potentiated by metformin still remains; therefore, it is necessary to determine the best combination strategies including metformin, the optimal dose to avoid adverse events, and lastly the potential biomarkers to implement it in the context of precision oncology.