Next Article in Journal
Identification of Reference Genes for Circadian Studies on Brain Microvessels and Choroid Plexus Samples Isolated from Rats
Previous Article in Journal
Post-Transcriptional Control of Mating-Type Gene Expression during Gametogenesis in Saccharomyces cerevisiae
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Biomolecules
Volume 11
Issue 8
10.3390/biom11081226
biomolecules-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Metformin and Risk of Malignant Brain Tumors in Patients with Type 2 Diabetes Mellitus

by
Chin-Hsiao Tseng
1,2,3
1
Department of Internal Medicine, National Taiwan University College of Medicine, Taipei 10051, Taiwan
2
Division of Endocrinology and Metabolism, Department of Internal Medicine, National Taiwan University Hospital, Taipei 10051, Taiwan
3
Division of Environmental Health and Occupational Medicine, National Health Research Institutes, Zhunan 350, Taiwan
Biomolecules 2021, 11(8), 1226; https://doi.org/10.3390/biom11081226
Submission received: 8 July 2021 / Revised: 9 August 2021 / Accepted: 12 August 2021 / Published: 17 August 2021
Browse Figure
Versions Notes

Abstract

:
The risk of malignant brain tumors associated with metformin use has rarely been investigated in humans. This retrospective cohort study investigated such an association. Patients with new-onset type 2 diabetes mellitus diagnosed from 1999 to 2005 in the nationwide database of Taiwan’s national health insurance were used to enroll study subjects. We first identified an unmatched cohort of 153,429 ever users and 16,222 never users of metformin. A cohort of 16,222 ever users and 16,222 never users matched on propensity score was then created from this unmatched cohort. All patients were followed up from 1 January 2006 until 31 December 2011. The incidence density was calculated and hazard ratios were derived from Cox regression incorporated with the inverse probability of treatment weighting using a propensity score. The results showed that 27 never users and 155 ever users developed malignant brain tumors in the unmatched cohort. The incidence rate was 37.11 per 100,000 person-years in never users and 21.39 per 100,000 person-years in ever users. The overall hazard ratio comparing ever users versus never users was 0.574 (95% confidence interval: 0.381–0.863). The respective hazard ratios comparing the first (<27.13 months), second (27.13–58.33 months), and third (>58.33 months) tertiles of cumulative duration of metformin therapy versus never users were 0.897 (0.567–1.421), 0.623 (0.395–0.984), and 0.316 (0.192–0.518). In the matched cohort, the overall hazard ratio was 0.317 (0.149–0.673) and the respective hazard ratios were 0.427 (0.129–1.412), 0.509 (0.196–1.322), and 0.087 (0.012–0.639) for the first, second, and third tertile of cumulative duration of metformin therapy. In conclusion, this study shows a risk reduction of malignant brain tumors associated with metformin use in a dose–response pattern. The risk reduction is more remarkable when metformin has been used for approximately 2–5 years.

1. Introduction

The incidence of malignant brain tumors (MBT) differs among different countries and glioblastoma is the most common adult primary MBT. A prospective cohort study that followed 8006 Japanese-American men living in Hawaii since 1965 until 1998 showed an incidence of glioblastoma of 6.2 per 100,000 person-years [1]. The age-adjusted incidence was reported to be 3.19 (during 2006–2010), 3.40 (during 2000–2008), 2.05 (during 1999–2003), 0.59 (during 2005), 3.69 (during 2005–2007), and 0.89 (during 2012–2013) per 100,000 population in the USA, Australia, UK, Korea, Greece, and Jordan, respectively [2]. In a recent statistical report derived from the Central Brain Tumor Registry of the United States (CBTRUS), the age-adjusted incidence rate of MBT during 2012–2016 was 7.08 per 100,000 population [3], suggesting an increase in age-adjusted incidence in the USA compared to the rate of 3.19 per 100,000 population during 2006–2010, which was derived from the same dataset of the CBTRUS [4]. In Taiwan, there is a gradual increase in MBT according to a report of the secular trends [5]. For the period from 1980–1984, 1985–1989, 1990–1994, 1995–1999, and 2000–2006, the age-standardized incidence rates per 100,000 population were 1.36, 1.74, 2.29, 2.47, and 2.54, respectively. Men had a higher incidence than women in each specific period, but the increasing trend was observed in both sexes [5].
There is no satisfactory treatment for MBT and the prognosis is poor, with a 5-year survival of <5% after diagnosis for glioblastoma [2]. Some genetic risk factors have been identified for MBT [6,7]. However, occupational and environmental exposures can also increase the risk [8]. These include sugar intake and carbon tetrachloride [1], radiation [9], immune factors and viral interaction [10], and air pollution with long-term exposure to PM2.5 [11].
Metformin is an oral antidiabetic drug that lowers blood glucose and improves insulin resistance. Its major mechanism of glucose-lowering action is by inhibiting hepatic gluconeogenesis through an inhibition of the mitochondrial respiratory chain complex 1. The reduction of hepatic energy status leads to an activation of the 5′ adenosine monophosphate-activated protein kinase (AMPK), a serine/threonine protein kinase [12,13,14]. Beyond glycemic control, metformin exerts a variety of pleiotropic benefits such as weight reduction, improvement of metabolic syndrome, reduction of cardiovascular morbidity, renoprotection, anti-infection, immune-modulation, anti-oxidation, anti-aging, and anti-cancer [14,15].
After oral intake, metformin can cross the blood–brain barrier and distribute to the brain tissues [16,17]. In recent years, the potential usefulness of metformin on the treatment of MBT has been investigated [18]. According to in vitro and in vivo studies, metformin may inhibit the growth of human glioblastoma cells and enhance therapeutic responses to chemotherapy (e.g., temozolomide) and radiotherapy [18,19,20]. Several clinical trials are being conducted to investigate its potential usefulness in the treatment of MBT but the outcomes remain unknown [17,21,22].
Since MBT has very poor prognosis without successful therapeutic modalities, prevention of the disease would be the best way to save lives and to reduce the disease load and economic burden. Whether metformin may exert a protective effect against MBT in humans has rarely been studied. To the best of our knowledge, there was only one previous matched case-control study, which used the UK-based Clinical Practice Research Datalink to investigate the risk of glioma associated with antidiabetic drugs [23]. This study concluded a lack of association between metformin use and glioma [23]. The aim of the present retrospective cohort study was to investigate whether the risk of MBT might be affected by metformin use and whether a dose–response relationship could be demonstrated.

2. Materials and Methods

2.1. National Health Insurance

This was a retrospective cohort study. The Taiwan’s National Health Insurance (NHI) is a nationwide healthcare system that has been implemented since 1 March 1995. The coverage rate of the NHI is very high and includes 99.6% of the whole population. All in-hospitals and 93% of all medical settings in Taiwan provide medical services under contracts with the Bureau of NHI. Computer files have to be submitted to the Bureau of NHI for reimbursement purpose and these files include information of diagnoses of diseases, prescriptions of drugs, and surgical procedures performed. Academic research proposals using the database can be approved after ethics review. The present study was reviewed and approved by the National Health Research Institutes (number NHIRD-102-175, approved on 5 September 2013). More detail descriptions of the database can be seen in previously published papers [24,25].

2.2. Study Population

Diagnoses of diseases in the database were coded by the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) during the whole study period. Accordingly, the codes used in the study for diabetes mellitus were 250 and the code of 191 was used for the diagnosis of MBT.
An unmatched original cohort and a matched cohort used for analyses in the study were created following the procedures shown in Figure 1 step-by-step. A total of 423,949 patients who had a first diagnosis of diabetes mellitus between 1999 and 2005 and had received outpatient prescriptions of antidiabetic drugs for ≥ 2 times were first identified. Ever users of metformin defined in the study were those who had been prescribed metformin as the first antidiabetic drug without receiving other antidiabetic drugs before the prescription of metformin (n = 183,837). Patients with the following conditions were then excluded: (1) patients who were diagnosed of type 1 diabetes mellitus (n = 2062), (2) patients who had missing data (n = 423), (3) patients who had suffered from any cancer before entry or within 6 months of diabetes diagnosis (n = 26,803), (4) patients who were aged < 25 years (n = 9275), (5) patients who were aged > 75 years (n = 27,296), and (6) patients who had been followed up for a short period of < 180 days (n = 4602). As a result, we identified 153,429 ever users and 16,222 never users of metformin and these patients were considered as the unmatched original cohort.
The propensity score (PS) was created by logistic regression that included all characteristics listed in Table 1 and the entry date as independent variables. The Greedy 8 to 1 digit match algorithm using the PS [26] was used to create a cohort of matched pairs of ever users and never users of metformin from the unmatched original cohort. This cohort was considered as the matched cohort.

2.3. Potential Confounders

Potential confounders derived from the database and their ICD-9-CM codes are shown in Table 1. These confounders were divided into six categories: (1) demographic and basic data, (2) major comorbidities, (3) diabetes-related complications, (4) antidiabetic drugs, (5) commonly encountered comorbidities, and (6) medications that are commonly used in diabetes patients. Occupations were classified into classes I to IV according to the Bureau of the NHI [27]. Class I included civil servants, teachers, employees of governmental or private businesses, professionals, and technicians. Class II included people without a specific employer, self-employed people, or seamen. Class III included farmers and fishermen. Class IV referred to low-income families supported by social welfare, or veterans.

2.4. Statistical Analyses

SAS statistical software, version 9.3 (SAS Institute, Cary, NC, USA) was used to conduct the statistical analyses. Statistical significance was set at a p value < 0.05.
The standardized difference between metformin ever users and never users was calculated for each covariate listed in Table 1. The standardized difference is recommended by Austin and Stuart as a test for balance diagnostics [28]. A threshold cutoff value of >10% in standardized difference was used as a potential indicator of imbalance in the variable, which might result in confounding from the variable [28].
The cumulative duration of metformin therapy expressed in months for each patient was calculated from the database. Patients were divided into three subgroups of metformin exposure according to the tertiles of cumulative duration to examine a potential dose–response relationship. The incidence density of MBT was calculated for never users and for users categorized as ever users and as users according to the tertiles of cumulative duration of metformin therapy. The case number of newly identified MBT during follow-up was the numerator of the incidence density. The denominator was expressed as person-years of follow-up. The date of 1 January 2006 was set as the starting date of follow-up. Follow-up ended up to 31 December 2011, at a time when any of the following events occurred first: a new-onset MBT, the date of death, or the date of the last reimbursement record available from the database.
Austin recommended the use of the inverse probability of treatment weighting (IPTW) method to reduce the potential confounding from the differences in covariates [29]. We therefore estimated hazard ratios and their 95% confidence intervals in the unmatched cohort and the matched cohort, respectively, by using Cox regression incorporated with the IPTW using the PS. Hazard ratios were estimated for ever users versus never users, and for ever users, divided by the tertiles of cumulative duration of metformin therapy versus never users.
Sensitivity analyses were then conducted in more restricted patients in the unmatched cohort. First, we excluded patients who had received any two consecutive metformin prescriptions spanning a period of > 4 months. As the prescription of drugs is not allowed to exceed 3 months at each time of prescription, as stipulated by the Bureau of the NHI, we might have excluded most patients with poor adherence when we excluded those patients who did not receive regular drug refills in the analyses. Second, we excluded patients who happened to be treated with incretins during follow-up to prevent potential confounding from the use of incretins (incretin-based therapies were not introduced into the market of Taiwan after the start of follow-up).

3. Results

The characteristics in never users of metformin and ever users of metformin and the standardized differences between these two groups of patients are shown in Table 1. In the unmatched original cohort, the values of standardized difference were >10% for age, dyslipidemia, obesity, nephropathy, eye diseases, insulin, meglitinide, acarbose, rosiglitazone, pioglitazone, statin, and fibrate, suggesting potential residual confounding from these variables. However, all values of standardized difference in the matched cohort were <10%, suggesting that the ever and never users of metformin in the matched cohort were well matched for the potential confounders.
Table 2 shows the incidence of MBT in different subgroups categorized according to metformin exposure and the hazard ratios comparing different subgroups of metformin exposure to a referent group of metformin never users. The overall hazard ratios suggested a significantly lower risk in ever users while compared to a referent group of never users. The findings derived from the unmatched cohort and the matched cohort were very similar and a dose–response pattern in terms of the cumulative duration of metformin therapy was observed in the tertile analyses. Metformin use for more than 2 years in the second tertile in the unmatched cohort and more than approximately 5 years in the third tertile in the matched cohort was significantly associated with a reduced risk.
Table 3 shows the sensitivity analyses after excluding patients who had not received regular refill of metformin and after excluding users of incretin-based therapies during follow-up. The results supported the findings in the main analyses shown in Table 2.

4. Discussion

4.1. Main Findings

The findings of the present study first provide evidence to support that, in patients with type 2 diabetes mellitus, metformin use would be significantly associated with an overall risk reduction of MBT, which could be similarly shown in either the unmatched cohort or the matched cohort in the main analyses (Table 2). A dose–response relationship could be seen in all analyses and a significant risk reduction could be seen when the cumulative duration of metformin therapy was more than 2–5 years (Table 2 and Table 3).

4.2. Limitations of an Early Study

The previous matched case-control (1:10) study, which used the UK-based Clinical Practice Research Datalink, included 2005 cases of glioma and 20,050 controls without glioma [23]. The investigators estimated odds ratios of 1.11 (0.59–2.12), 1.42 (0.81–2.47), and 0.72 (0.38–1.39) for patients who received numbers of metformin prescriptions of 1–9, 10–29 and ≥ 30, respectively, in comparison to cases and controls without metformin use [23]. They concluded a lack of significant risk association between MBT and metformin use [23]. However, there are probably some limitations that might have led to the misinterpretation of a lack of effect associated with metformin use in this previous study. First, the investigators included diabetes patients and non-diabetic people in the study and there were only 96 diabetes patients (4.79%) in the cases with glioma (among them: 57 metformin users) and 1240 diabetes patients (6.18%) in the control group without glioma (among them: 716 metformin users) [23]. The numbers of diabetes patients in the cases and controls were actually very small. Second, when they compared the odds ratio for metformin use in each specific subgroup of the number of metformin prescriptions versus non-metformin use, they actually compared metformin users in the diabetes patients versus a group of non-metformin users that was composed mainly of non-diabetic people. Even though the diabetes status had been additionally adjusted for in the models, the small numbers of glioma cases (14, 23, and 20 cases) and the small numbers of metformin users (184, 239, and 350) in the categories of 1–9, 10–29, and ≥ 30 prescriptions of metformin, respectively, might have led to biased estimates because of a lack of sufficient power in the assessment of risk association. Third, in their model, which considered only diabetes patients matched on diabetes duration and A1C level, there were only 86 cases of glioma and 598 controls without glioma. The odds ratios for the numbers of metformin prescriptions of 1–9, 10–29, and ≥30 were 1.07 (0.47–2.40), 1.06 (0.53–2.11), and 0.58 (0.24–1.44), respectively. There seemed to be a neutral effect in the first two subgroups with a lower exposure to metformin but a lower risk could be seen (though not statistically significant) in the third subgroup that had been exposed to a higher cumulative dose of metformin. Similarly, the small number in each diabetes subgroup might have led to a biased conclusion because of a lack of statistical power.

4.3. Potential Mechanisms

The mechanisms of the potential protective effect of metformin on MBT remain unknown, but some biological effects of metformin can explain such an observation. MBT cells are dependent on glucose metabolism (a phenomenon known as the Warburg effect) and the mammalian target of rapamycin (mTOR) signaling to support their proliferation and growth [18,30]. Metformin inhibits the mitochondrial complex 1 of electron transport and reduces energy supply to cancer cells [12,18]. Metformin has been shown to inhibit glioblastoma cell growth and induce cell cycle arrest, autophagy, and apoptosis in in vitro studies. These were parallel to an activation of the AMPK and an inhibition of the mTOR pathway and were dependent on genetic and mutational backgrounds [31]. Metformin may also target glioma stem cells, leading to cell cycle arrest and mitochondria-dependent apoptosis [32]. AMPK activation by metformin may also activate the transcriptional activity of p53, a tumor suppressor gene, in MBT cells [18,30]. Lipid peroxidation plays an important role in the development of MBT [33] and MBT cell growth can also be inhibited by reduced fatty acid synthesis and beta-oxidation [30]. Long-term treatment with metformin does inhibit fatty acid synthase and decrease the expression of genes involved in fatty acid oxidation [12]. Leptin, an adipokine that plays important roles in energy balance in the brain, promotes angiogenesis and cell proliferation, survival, and migration via signaling pathways involving the Janus kinase family, signal transducers and activators of transcription, mitogen-activated protein kinase, phosphoinositide-3-kinase, and mTOR, etc. [21]. Metformin can cross the blood–brain barrier and correct these dysregulated pathways of leptin via the activation of AMPK [21]. Ionizing radiation has been linked to the development of MBT in adults and children [9,34]. It is interesting that metformin exerts protection against radiation-induced damage to the lung [35] and skin [36] in animals via the inhibition of inflammatory cytokines.

4.4. Implications

First, many ongoing clinical trials are being conducted to investigate the efficacy of metformin as a therapeutic agent for glioblastoma [17,21,22]. The findings of the present study provide some useful clinical information. The requirement of a prolonged use of metformin for more than 2–5 years for a significant preventive effect to be observed (Table 2 and Table 3) implies that a higher dose or even longer duration of therapy may be necessary if metformin is used as a therapeutic agent. However, because of the highly malignant characteristics of most MBTs [2], the usefulness of metformin as a therapeutic agent may be questionable and requires more in-depth investigation.
Second, our previous observational studies conducted in patients with type 2 diabetes mellitus in Taiwan by using the nationwide NHI database suggest that metformin use is associated with a lower risk of various types of cancer, including lung cancer [24], colorectal cancer [25,27], breast cancer [37], thyroid cancer [38], bladder cancer [39], prostate cancer [40], endometrial cancer [41], ovarian cancer [42], cervical cancer [43], kidney cancer [44], oral cancer [45], gastric cancer [46], esophageal cancer [47], nasopharyngeal cancer [48], skin cancer [49], pancreatic cancer [50], hepatocellular cancer [51], biliary tract cancer [52], non-Hodgkin’s lymphoma [53], and bone cancer [54]. These findings suggest that the anti-cancer effects of metformin may involve some common pathophysiological pathways in cancer development, probably targeting many of the hallmarks of cancer [55].
Third, in recent years, some novel antidiabetic drugs such as sodium–glucose cotransporter-2 inhibitors and glucagon-like peptide-1 receptor agonists have been shown to provide cardiovascular and renal protection in patients with type 2 diabetes mellitus [56]. These newer drugs have challenged the positioning of metformin as the first-line therapeutic drug recommended by the American Diabetes Association and the European Association of the Study of Diabetes since 2008 [57]. However, these newer drugs are expensive and their long-term safety, especially with regards to cancer risk, remains unknown. As pointed out by a recent article discussed by Baker et al., up to now, “there are no data to suggest that metformin should not be initiated soon after the diagnosis of diabetes”, even in the era of these novel antidiabetic drugs [56]. Our previous pharmaco-epidemiological studies conducted in Taiwan suggest that metformin not only reduces cancer risk, but also the risk of various non-malignant diseases such as hypertension [58], heart failure [59], atrial fibrillation [60], chronic obstructive pulmonary disease [61], pulmonary tuberculosis infection [62], Helicobacter pylori infection [63], varicose veins [64], acute appendicitis [65], hemorrhoids [66], dementia [67,68], nodular goiter [69], uterine leiomyoma [70], osteoporosis/vertebral fracture [71], and inflammatory bowel disease [72]. We also found that metformin ever users have a lower risk of total mortality than never users of metformin, with an estimated multivariate-adjusted hazard ratio of 0.67 (95% confidence interval: 0.64–0.69) [72]. Therefore, besides the low cost and minimal side effects, the general beneficial effects of metformin on malignant and non-malignant human diseases and on total mortality observed in our previous studies provide a good rationale for using metformin as the first-line therapeutic drug for type 2 diabetes mellitus.

4.5. Strengths

Some methodological limitations are commonly seen in pharmaco-epidemiological studies that use big administrative databases to examine the potential clinical outcomes related to medications. In the present study, we carefully addressed these potential limitations to avoid selection bias, prevalent user bias, immortal time bias, and confounding by indication.
The large representative sample derived from the nationwide NHI database is a strong cohort because the database covers 99.6% of the Taiwan’s population. Selection bias could be avoided and the findings can well be generalized to the whole population of Taiwan. To prevent the potential occurrence of prevalent user bias, we purposely included only patients who were newly diagnosed of diabetes mellitus and who were new users of metformin into the study (Figure 1).
Inappropriately assigning the treatment status and follow-up time in the calculation of the follow-up period in the compared subgroups can induce immortal time bias. In the present study, misclassification of the treatment status was not likely by using the universal healthcare system, which keeps all prescription information since 1995, and by enrolling patients who had received ≥2 times of documented prescriptions of antidiabetic drugs (Figure 1). While calculating the follow-up time, we deliberately excluded the immortal time that could happen when the patients were handled with a non-pharmacological approach after diabetes diagnosis (i.e., the time in between diabetes diagnosis and the first antidiabetic drugs prescribed to the patients). Furthermore, we excluded the immortal time that could happen during the initial period of follow-up (i.e., <180 days) (Figure 1). It is worth mentioning that the immortal time that happens within the waiting period between the date of hospital discharge and the dispensing of the drugs prescribed at discharge would not happen in Taiwan. This is because of the fact that in our NHI healthcare system, all discharge drugs of a patient can be obtained immediately from the hospital on the date of the patient’s discharge.
Confounding by indication refers to an association of a risk factor with the indication of a medication under investigation. This can be handled by using a matched cohort of users and non-users balancing in potential confounders [28] and by modeling with Cox regression incorporated with IPTW using PS [29]. The consistency of the findings (Table 2 and Table 3) strengthened a preventive effect of metformin on MBT.
Additionally, there are some other strengths. First, to minimize reverse causality, patients with a diagnosis of MBT within 6 months of diabetes diagnoses had been excluded (Figure 1). Second, the use of pre-existing medical records in the NHI database could reduce the bias resulting from self-reporting. Third, detection bias because of different socioeconomic status could be much reduced because in the NHI healthcare system, the copayments are low and most of them can be waived for low-income patients, veterans, and patients who receive drug refills for chronic diseases.

4.6. Limitations

It is acknowledged that there are some study limitations. First, this is a pharmaco-epidemiological study and the basic mechanisms of metformin’s protective effects on the tumorigenesis of MBT could not be investigated.
Second, malignant cells of brain tumors may come from different brain cell types and may have different gene mutations and specifications of metabolic pathways (e.g., glycolytic, glutaminolytic, or oxidative phenotypes) [73]. However, we did not have histopathological data of MBT for disease confirmation and could only use the ICD-9-CM code as a diagnostic tool. The ICD-9-CM code of 191 for MBT diagnosis does not provide detailed information on the histopathology of MBT as classified by the World Health Organization [74].
Third, it should also be noted that the major histopathological types of MBT may differ between adults and children [75,76]. Medulloblastoma is the most commonly seen embryonal tumor in children (approximately 25% of all brain malignancies in children), but it represents < 1% in adults (mainly in young adults < 40 years old) [76,77]. Neuroblastoma is another embryonic cancer involving the sympathetic nervous system that is mainly seen in infants and young children [78]. As metformin is only approved for the treatment of type 2 diabetes mellitus, which is not commonly seen in children, we have excluded patients aged < 25 years in the study (Figure 1). Therefore, the findings of the present study should better not be applied to childhood MBT such as medulloblastoma and neuroblastoma before additional research and confirmation. Neither should the findings be applied to other rare forms of MBT such as ependymoma and sarcoma because they only represent approximately 3% and 2% of MBT, respectively, in Taiwan [79].
Fourth, approximately 80% of MBT arises from glia cells [80] but secondary glioblastomas with genetic mutations in isocitrate dehydrogenase, TP53 or ATRX, etc. are rare [81]. Therefore, primary glioblastoma should be the major category of MBT observed in this study. Metabolic reprogramming has been found to be an important cancer hallmark [82,83] and heterogeneity in the preference of bioenergetic pathways exists in glioblastomas, which can lead to their discrepant sensitivity to metformin [73]. In a recent in vitro study that used three glioblastoma stem cells (GBM18, GBM27, and GBM38) and one human glioblastoma cell line (U87MG) [73], GBM18 seemed to be the most sensitive cell line and GBM27 the most resistant cell line to metformin [73]. GBM18 is characterized by a Warburg-like (glycolytic) metabolism but GBM27 is characterized by a highly oxidative metabolism (up to 50%) with a slower proliferation rate [73]. Such a discrepancy observed in different cell lines was also relative to the effects of metformin on the activation of AMPK, resulting in the inhibition of mTOR [73]. This in vitro study also showed that metformin would affect the survival of normal stem cells less while administered to glioblastoma cancer cells of Warburg-like phenotypes [73]. Therefore, it is reasonable to expect that the effect of metformin on the prevention of MBT might not be the same to all tumors classified under the category of glioblastoma by the World Health Organization [74]. We were not able to answer questions related to such different bioenergetic phenotypes and more in-depth investigation should be tailored to the bioenergetic phenotypes of the tumor cells in future observational studies or interventional clinical trials.
Finally, unmeasured confounders could never be adjusted for and their impacts remained unknown in the study. These potential confounders may include the use of mobile phones, radiation therapy, air pollution, household conditions, education levels, lifestyle, smoking, alcohol drinking, nutritional status, dietary pattern, anthropometric factors, biochemistry, immune profiles, family history, and genetic parameters. However, because a confounder has to be correlated with the exposure (metformin use) and the disease (MBT) at the same time and it must not be in the causal pathway between them [84], there seemed to be no strong evidence to suggest that these unmeasured factors would be correlated with metformin (exposure).

5. Conclusions

In summary, this observational retrospective cohort study is the first to suggest a preventive effect exerted by metformin on MBT incidence in patients with type 2 diabetes mellitus. Confirmation with additional consideration of potential confounders, histopathological data, bioenergetic phenotypes, and/or genetic biomarkers is required. Metformin is cheap and safe and its usefulness as a preventive drug or a chemotherapeutic adjuvant to MBT in either diabetes patients or non-diabetic people is worthy of more extensive investigation with prospective cohort designs or randomized controlled clinical trials. Nanotechnology is being applied to develop metformin-derived carbon dots that can more readily cross the blood–brain barrier and accumulate more selectively inside the mitochondria of cancer cells [85]. The application of this novel agent in the treatment of MBT and the tailoring treatment based on metabolic phenotypes of the cancer should provide more promising effects.

Funding

The study was supported by the Ministry of Science and Technology (MOST 107-2221-E-002-129-MY3) of Taiwan and the Yee Fong Charity Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Institutional Review Board Statement

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Review Board of the National Health Research Institutes (number NHIRD-102-175, approved on 5 September 2013).

Informed Consent Statement

Personal information was not available in the datasets and informed consent from the participants was not required according to local regulations.

Data Availability Statement

The datasets presented in this article are not readily available because public availability of the dataset is restricted by local regulations to protect privacy.

Conflicts of Interest

The author declares no conflict of interest.

References

  1. Nelson, J.S.; Burchfiel, C.M.; Fekedulegn, D.; Andrew, M.E. Potential risk factors for incident glioblastoma multiforme: The Honolulu Heart Program and Honolulu-Asia Aging Study. J. Neurooncol. 2012, 109, 315–321. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  2. Tamimi, A.F.; Juweid, M. Epidemiology and Outcome of Glioblastoma; De Vleeschouwer, S., Ed.; Exon Publications: Brisbane, Australia, 2017; Chapter 8. Available online: https://www.ncbi.nlm.nih.gov/books/NBK470003/ (accessed on 14 August 2021).
  3. Ostrom, Q.T.; Cioffi, G.; Gittleman, H.; Patil, N.; Waite, K.; Kruchko, C.; Barnholtz-Sloan, J.S. CBTRUS Statistical Report: Primary brain and other central nervous system tumors diagnosed in the United States in 2012-2016. Neuro. Oncol. 2019, 21, v1–v100. [Google Scholar] [CrossRef] [PubMed]
  4. Ostrom, Q.T.; Gittleman, H.; Farah, P.; Ondracek, A.; Chen, Y.; Wolinsky, Y.; Stroup, N.E.; Kruchko, C.; Barnholtz-Sloan, J.S. CBTRUS statistical report: Primary brain and central nervous system tumors diagnosed in the United States in 2006-2010. Neuro. Oncol. 2013, 15, ii1–ii56. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  5. Chiang, C.J.; Chen, Y.C.; Chen, C.J.; You, S.L.; Lai, M.S.; Taiwan Cancer Registry Task Force. Cancer trends in Taiwan. Jpn J. Clin. Oncol. 2010, 40, 897–904. [Google Scholar] [CrossRef] [PubMed]
  6. Louis, D.N.; Perry, A.; Reifenberger, G.; von Deimling, A.; Figarella-Branger, D.; Cavenee, W.K.; Ohgaki, H.; Wiestler, O.D.; Kleihues, P.; Ellison, D.W. The 2016 World Health Organization classification of tumors of the central nervous system: A summary. Acta Neuropathol. 2016, 131, 803–820. [Google Scholar] [CrossRef] [Green Version]
  7. Komori, T. The 2016 WHO classification of tumours of the central nervous system: The major points of revision. Neurol. Med.-Chir. 2017, 57, 301–311. [Google Scholar] [CrossRef] [Green Version]
  8. Gomes, J.; Al Zayadi, A.; Guzman, A. Occupational and environmental risk factors of adult primary brain cancers: A systematic review. Int. J. Occup. Environ. Med. 2011, 2, 82–111. [Google Scholar] [PubMed]
  9. Ostrom, Q.T.; Gittleman, H.; Stetson, L.; Virk, S.M.; Barnholtz-Sloan, J.S. Epidemiology of gliomas. Cancer Treat. Res. 2015, 163, 1–14. [Google Scholar]
  10. Johnson, K.J.; Hainfellner, J.A.; Lau, C.C.; Scheurer, M.E.; Woehrer, A.; Wiemels, J. Immune factors and viral interactions in brain cancer etiology and outcomes, the 2016 Brain Tumor Epidemiology Consortium Meeting report. Clin. Neuropathol. 2016, 35, 280–286. [Google Scholar] [CrossRef] [Green Version]
  11. Andersen, Z.J.; Pedersen, M.; Weinmayr, G.; Stafoggia, M.; Galassi, C.; Jørgensen, J.T.; Sommar, J.N.; Forsberg, B.; Olsson, D.; Oftedal, B.; et al. Long-term exposure to ambient air pollution and incidence of brain tumor: The European Study of Cohorts for Air Pollution Effects (ESCAPE). Neuro. Oncol. 2018, 20, 420–432. [Google Scholar] [CrossRef] [Green Version]
  12. Maniar, K.; Moideen, A.; Mittal, A.; Patil, A.; Chakrabarti, A.; Banerjee, D. A story of metformin-butyrate synergism to control various pathological conditions as a consequence of gut microbiome modification: Genesis of a wonder drug? Pharmacol. Res. 2017, 117, 103–128. [Google Scholar] [CrossRef] [PubMed]
  13. Rena, G.; Hardie, D.G.; Pearson, E.R. The mechanisms of action of metformin. Diabetologia 2017, 60, 1577–1585. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Mbara, K.C.; Mofo Mato, P.E.; Driver, C.; Nzuza, S.; Mkhombo, N.T.; Gcwensa, S.K.; Mcobothi, E.N.; Owira, P.M. Metformin turns 62 in pharmacotherapy: Emergence of non-glycaemic effects and potential novel therapeutic applications. Eur. J. Pharmacol. 2021, 898. [Google Scholar] [CrossRef] [PubMed]
  15. Yu, X.; Mao, W.; Zhai, Y.; Tong, C.; Liu, M.; Ma, L.; Yu, X.; Li, S. Anti-tumor activity of metformin: From metabolic and epigenetic perspectives. Oncotarget 2017, 8, 5619–5628. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  16. Łabuzek, K.; Suchy, D.; Gabryel, B.; Bielecka, A.; Liber, S.; Okopień, B. Quantification of metformin by the HPLC method in brain regions, cerebrospinal fluid and plasma of rats treated with lipopolysaccharide. Pharmacol. Rep. 2010, 62, 956–965. [Google Scholar] [CrossRef]
  17. El-Arabey, A.A.; Abdalla, M.; Ali Eltayb, W. Metformin: Ongoing journey with superdrug revolution. Adv. Pharm. Bull. 2019, 9, 1–4. [Google Scholar] [CrossRef] [PubMed]
  18. Takhwifa, F.; Aninditha, T.; Setiawan, H.; Sauriasari, R. The potential of metformin as an antineoplastic in brain tumors: A systematic review. Heliyon 2021, 7. [Google Scholar] [CrossRef]
  19. Seliger, C.; Meyer, A.L.; Renner, K.; Leidgens, V.; Moeckel, S.; Jachnik, B.; Dettmer, K.; Tischler, U.; Gerthofer, V.; Rauer, L.; et al. Metformin inhibits proliferation and migration of glioblastoma cells independently of TGF-β2. Cell Cycle 2016, 15, 1755–1766. [Google Scholar] [CrossRef] [Green Version]
  20. Kolesnik, D.L.; Pyaskovskaya, O.N.; Yurchenko, O.V.; Solyanik, G.I. Metformin enhances antitumor action of sodium dichloroacetate against glioma C6. Exp. Oncol. 2019, 41, 123–129. [Google Scholar] [CrossRef]
  21. Kinfe, T.M.; Stadlbauer, A.; Bozhkov, Y.; Kremenevski, N.; Brandner, S.; Buchfelder, M.; Chaudhry, S.R. The diagnostic and therapeutic role of leptin and its receptor ObR in glioblastoma multiforme. Cancers 2020, 12, 3691. [Google Scholar] [CrossRef] [PubMed]
  22. Porper, K.; Shpatz, Y.; Plotkin, L.; Pechthold, R.G.; Talianski, A.; Champ, C.E.; Furman, O.; Shimoni-Sebag, A.; Symon, Z.; Amit, U.; et al. A phase I clinical trial of dose-escalated metabolic therapy combined with concomitant radiation therapy in high-grade glioma. J. Neurooncol. 2021, 153, 487–496. [Google Scholar] [CrossRef] [PubMed]
  23. Seliger, C.; Ricci, C.; Meier, C.R.; Bodmer, M.; Jick, S.S.; Bogdahn, U.; Hau, P.; Leitzmann, M.F. Diabetes, use of antidiabetic drugs, and the risk of glioma. Neuro. Oncol. 2016, 18, 340–349. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  24. Tseng, C.H. Metformin and lung cancer risk in patients with type 2 diabetes mellitus. Oncotarget 2017, 8, 41132–41142. [Google Scholar] [CrossRef]
  25. Tseng, C.H. Metformin is associated with a lower risk of colorectal cancer in Taiwanese patients with type 2 diabetes: A retrospective cohort analysis. Diabetes Metab. 2017, 43, 438–445. [Google Scholar] [CrossRef]
  26. Parsons, L.S. Performing a 1:N Case-Control Match on Propensity Score. Available online: http://www.google.com.tw/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CBsQFjAAahUKEwibi7HllcnIAhUDoJQKHVeZA9A&url=http%3A%2F%2Fwww2.sas.com%2Fproceedings%2Fsugi29%2F165-29.pdf&usg=AFQjCNFOHGWYu8E8Bn4-Bo1TUiJKtT987Q (accessed on 2 July 2021).
  27. Tseng, C.H. Diabetes, metformin use, and colon cancer: A population-based cohort study in Taiwan. Eur. J. Endocrinol. 2012, 167, 409–416. [Google Scholar] [CrossRef] [Green Version]
  28. Austin, P.C.; Stuart, E.A. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat. Med. 2015, 34, 3661–3679. [Google Scholar] [CrossRef]
  29. Austin, P.C. The performance of different propensity score methods for estimating marginal hazard ratios. Stat. Med. 2013, 32, 2837–2849. [Google Scholar] [CrossRef] [Green Version]
  30. Strickland, M.; Stoll, E.A. Metabolic reprogramming in glioma. Front. Cell Dev. Biol. 2017, 5. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  31. Sesen, J.; Dahan, P.; Scotland, S.J.; Saland, E.; Dang, V.T.; Lemarié, A.; Tyler, B.M.; Brem, H.; Toulas, C.; Cohen-Jonathan Moyal, E.; et al. Metformin inhibits growth of human glioblastoma cells and enhances therapeutic response. PLoS ONE 2015, 10. [Google Scholar] [CrossRef] [Green Version]
  32. Najbauer, J.; Kraljik, N.; Németh, P. Glioma stem cells: Markers, hallmarks and therapeutic targeting by metformin. Pathol. Oncol. Res. 2014, 20, 789–797. [Google Scholar] [CrossRef] [PubMed]
  33. Jaganjac, M.; Cindrić, M.; Jakovčević, A.; Žarković, K.; Žarković, N. Lipid peroxidation in brain tumors. Neurochem. Int. 2021, 149. [Google Scholar] [CrossRef]
  34. Ostrom, Q.T.; Adel Fahmideh, M.; Cote, D.J.; Muskens, I.S.; Schraw, J.M.; Scheurer, M.E.; Bondy, M.L. Risk factors for childhood and adult primary brain tumors. Neuro. Oncol. 2019, 21, 1357–1375. [Google Scholar] [CrossRef] [PubMed]
  35. Azmoonfar, R.; Amini, P.; Saffar, H.; Rezapoor, S.; Motevaseli, E.; Cheki, M.; Yahyapour, R.; Farhood, B.; Nouruzi, F.; Khodamoradi, E.; et al. Metformin protects against radiation-induced pneumonitis and fibrosis and attenuates upregulation of dual oxidase genes expression. Adv. Pharm. Bull. 2018, 8, 697–704. [Google Scholar] [CrossRef]
  36. Xiao, T.; Chen, Y.; Song, C.; Xu, S.; Lin, S.; Li, M.; Chen, X.; Gu, H. Possible treatment for UVB-induced skin injury: Anti-inflammatory and cytoprotective role of metformin in UVB-irradiated keratinocytes. J. Dermatol. Sci. 2021, 102, 25–35. [Google Scholar] [CrossRef] [PubMed]
  37. Tseng, C.H. Metformin may reduce breast cancer risk in Taiwanese women with type 2 diabetes. Breast Cancer Res. Treat. 2014, 145, 785–790. [Google Scholar] [CrossRef]
  38. Tseng, C.H. Metformin reduces thyroid cancer risk in Taiwanese patients with type 2 diabetes. PLoS ONE 2014, 9. [Google Scholar] [CrossRef]
  39. Tseng, C.H. Metformin may reduce bladder cancer risk in Taiwanese patients with type 2 diabetes. Acta Diabetol. 2014, 51, 295–303. [Google Scholar] [CrossRef] [PubMed]
  40. Tseng, C.H. Metformin significantly reduces incident prostate cancer risk in Taiwanese men with type 2 diabetes mellitus. Eur. J. Cancer 2014, 50, 2831–2837. [Google Scholar] [CrossRef] [PubMed]
  41. Tseng, C.H. Metformin and endometrial cancer risk in Chinese women with type 2 diabetes mellitus in Taiwan. Gynecol. Oncol. 2015, 138, 147–153. [Google Scholar] [CrossRef] [PubMed]
  42. Tseng, C.H. Metformin reduces ovarian cancer risk in Taiwanese women with type 2 diabetes mellitus. Diabetes Metab. Res. Rev. 2015, 31, 619–626. [Google Scholar] [CrossRef]
  43. Tseng, C.H. Metformin use and cervical cancer risk in female patients with type 2 diabetes. Oncotarget 2016, 7, 59548–59555. [Google Scholar] [CrossRef] [Green Version]
  44. Tseng, C.H. Use of metformin and risk of kidney cancer in patients with type 2 diabetes. Eur. J. Cancer 2016, 52, 19–25. [Google Scholar] [CrossRef]
  45. Tseng, C.H. Metformin may reduce oral cancer risk in patients with type 2 diabetes. Oncotarget 2016, 7, 2000–2008. [Google Scholar] [CrossRef] [Green Version]
  46. Tseng, C.H. Metformin reduces gastric cancer risk in patients with type 2 diabetes mellitus. Aging (Albany NY). 2016, 8, 1636–1649. [Google Scholar] [CrossRef] [Green Version]
  47. Tseng, C.H. Metformin and esophageal cancer risk in Taiwanese patients with type 2 diabetes mellitus. Oncotarget 2017, 8, 18802–18810. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  48. Tseng, C.H. Metformin and risk of developing nasopharyngeal cancer in patients with type 2 diabetes mellitus. Metabolism 2018, 85, 223–226. [Google Scholar] [CrossRef] [PubMed]
  49. Tseng, C.H. Metformin is associated with decreased skin cancer risk in Taiwanese patients with type 2 diabetes. J. Am. Acad. Dermatol. 2018, 78, 694–700. [Google Scholar] [CrossRef]
  50. Tseng, C.H. Metformin and pancreatic cancer risk in patients with type 2 diabetes. Pancreas 2018, 47, e57–e59. [Google Scholar] [CrossRef]
  51. Tseng, C.H. Metformin and risk of hepatocellular carcinoma in patients with type 2 diabetes. Liver Int. 2018, 38, 2018–2027. [Google Scholar] [CrossRef]
  52. Tseng, C.H. Metformin and biliary tract cancer in patients with type 2 diabetes. Front. Oncol. 2020, 10. [Google Scholar] [CrossRef] [PubMed]
  53. Tseng, C.H. Metformin is associated with a lower risk of non-Hodgkin lymphoma in patients with type 2 diabetes. Diabetes Metab. 2019, 45, 458–464. [Google Scholar] [CrossRef]
  54. Tseng, C.H. Metformin and primary bone cancer risk in Taiwanese patients with type 2 diabetes mellitus. Bone 2021, 151. [Google Scholar] [CrossRef]
  55. Hanahan, D.; Weinberg, R.A. Hallmarks of cancer: The next generation. Cell 2011, 144, 646–674. [Google Scholar] [CrossRef] [Green Version]
  56. Baker, C.; Retzik-Stahr, C.; Singh, V.; Plomondon, R.; Anderson, V.; Rasouli, N. Should metformin remain the first-line therapy for treatment of type 2 diabetes? Ther. Adv. Endocrinol. Metab. 2021, 12. [Google Scholar] [CrossRef]
  57. Nathan, D.M.; Buse, J.B.; Davidson, M.B.; Ferrannini, E.; Holman, R.R.; Sherwin, R.; Zinman, B. Medical management of hyperglycemia in type 2 diabetes: A consensus algorithm for the initiation and adjustment of therapy. Diabetes Care. 2008, 31, 1–11. [Google Scholar] [CrossRef] [Green Version]
  58. Tseng, C.H. Metformin and risk of hypertension in Taiwanese patients with type 2 diabetes mellitus. J. Am. Heart Assoc. 2018, 6. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  59. Tseng, C.H. Metformin use is associated with a lower risk of hospitalization for heart failure in patients with type 2 diabetes mellitus: A retrospective cohort analysis. J. Am. Heart Assoc. 2019, 8. [Google Scholar] [CrossRef] [PubMed]
  60. Tseng, C.H. Metformin use is associated with a lower incidence of hospitalization for atrial fibrillation in patients with type 2 diabetes mellitus. Front. Med. 2021, 7. [Google Scholar] [CrossRef]
  61. Tseng, C.H. Metformin and risk of chronic obstructive pulmonary disease in diabetes patients. Diabetes Metab. 2019, 45, 184–190. [Google Scholar] [CrossRef]
  62. Tseng, C.H. Metformin decreases risk of tuberculosis infection in type 2 diabetes patients. J. Clin. Med. 2018, 7, 264. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  63. Tseng, C.H. Metformin and Helicobacter pylori infection in patients with type 2 diabetes. Diabetes Care. 2018, 41, e42–e43. [Google Scholar] [CrossRef] [Green Version]
  64. Tseng, C.H. Metformin reduces risk of varicose veins in patients with type 2 diabetes. Diabetes Metab. Res. Rev. 2020, 36. [Google Scholar] [CrossRef] [PubMed]
  65. Tseng, C.H. Metformin use is associated with a reduced risk of acute appendicitis in Taiwanese patients with type 2 diabetes mellitus. Sci. Rep. 2021, 11. [Google Scholar] [CrossRef] [PubMed]
  66. Tseng, C.H. Chronic metformin therapy is associated with a lower risk of hemorrhoid in patients with type 2 diabetes mellitus. Front. Pharmacol. 2021, 11. [Google Scholar] [CrossRef]
  67. Tseng, C.H. Metformin and the risk of dementia in type 2 diabetes patients. Aging Dis. 2019, 10, 37–48. [Google Scholar]
  68. Tseng, C.H. Dementia risk in type 2 diabetes patients: Acarbose use and its joint effects with metformin and pioglitazone. Aging Dis. 2020, 11, 658–667. [Google Scholar] [CrossRef]
  69. Tseng, C.H. Metformin reduces risk of benign nodular goiter in patients with type 2 diabetes mellitus. Eur. J. Endocrinol. 2019, 180, 367–374. [Google Scholar] [CrossRef]
  70. Tseng, C.H. Metformin use is associated with a lower risk of uterine leiomyoma in female type 2 diabetes patients. Ther Adv. Endocrinol. Metab. 2019, 10. [Google Scholar] [CrossRef]
  71. Tseng, C.H. Metformin use is associated with a lower risk of osteoporosis/vertebral fracture in Taiwanese patients with type 2 diabetes mellitus. Eur. J. Endocrinol. 2021, 184, 299–310. [Google Scholar] [CrossRef]
  72. Tseng, C.H. Metformin use is associated with a lower risk of inflammatory bowel disease in patients with type 2 diabetes mellitus. J. Crohns. Colitis. 2021, 15, 64–73. [Google Scholar] [CrossRef] [PubMed]
  73. Duraj, T.; García-Romero, N.; Carrión-Navarro, J.; Madurga, R.; Mendivil, A.O.; Prat-Acin, R.; Garcia-Cañamaque, L.; Ayuso-Sacido, A. Beyond the Warburg effect: Oxidative and glycolytic phenotypes coexist within the metabolic heterogeneity of glioblastoma. Cells 2021, 10, 202. [Google Scholar] [CrossRef]
  74. Louis, D.N.; Perry, A.; Wesseling, P.; Brat, D.J.; Cree, I.A.; Figarella-Branger, D.; Hawkins, C.; Ng, H.K.; Pfister, S.M.; Reifenberger, G.; et al. The 2021 WHO Classification of Tumors of the Central Nervous System: A summary. Neuro. Oncol. 2021, 23, 1231–1251. [Google Scholar] [CrossRef]
  75. Fangusaro, J.; Bandopadhayay, P. Advances in the classification and treatment of pediatric brain tumors. Curr. Opin. Pediatr. 2021, 33, 26–32. [Google Scholar] [CrossRef]
  76. Merchant, T.E.; Pollack, I.F.; Loeffler, J.S. Brain tumors across the age spectrum: Biology, therapy, and late effects. Semin. Radiat. Oncol. 2010, 20, 58–66. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  77. Majd, N.; Penas-Prado, M. Updates on management of adult medulloblastoma. Curr. Treat. Options Oncol. 2019, 20. [Google Scholar] [CrossRef]
  78. Fetahu, I.S.; Taschner-Mandl, S. Neuroblastoma and the epigenome. Cancer Metastasis Rev. 2021, 40, 173–189. [Google Scholar] [CrossRef]
  79. Health Promotion Administration, Ministry of Health and Welfare, Taiwan. Cancer Registry Annual Report 2018. Taiwan. Available online: https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=269&pid=13498 (accessed on 4 August 2021).
  80. Alghamri, M.S.; McClellan, B.L.; Hartlage, M.S.; Haase, S.; Faisal, S.M.; Thalla, R.; Dabaja, A.; Banerjee, K.; Carney, S.V.; Mujeeb, A.A.; et al. Targeting neuroinflammation in brain cancer: Uncovering mechanisms, pharmacological targets, and neuropharmaceutical developments. Front. Pharmacol. 2021, 12. [Google Scholar] [CrossRef]
  81. Ohgaki, H.; Kleihues, P. The definition of primary and secondary glioblastoma. Clin. Cancer Res. 2013, 19, 764–772. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  82. Ward, P.S.; Thompson, C.B. Metabolic reprogramming: A cancer hallmark even Warburg did not anticipate. Cancer Cell. 2012, 21, 297–308. [Google Scholar] [CrossRef] [Green Version]
  83. Cantor, J.R.; Sabatini, D.M. Cancer cell metabolism: One hallmark, many faces. Cancer Discov. 2012, 2, 881–898. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  84. Van Stralen, K.J.; Dekker, F.W.; Zoccali, C.; Jager, K.J. Confounding. Nephron Clin. Pract. 2010, 116. [Google Scholar] [CrossRef] [PubMed]
  85. Kirbas Cilingir, E.; Seven, E.S.; Zhou, Y.; Walters, B.M.; Mintz, K.J.; Pandey, R.R.; Wikramanayake, A.H.; Chusuei, C.C.; Vanni, S.; Graham, R.M.; et al. Metformin derived carbon dots: Highly biocompatible fluorescent nanomaterials as mitochondrial targeting and blood-brain barrier penetrating biomarkers. J. Colloid Interface Sci. 2021, 592, 485–497. [Google Scholar] [CrossRef] [PubMed]
Biomolecules 11 01226 g001 550
Figure 1. The step-by-step procedures followed to create an unmatched original cohort and a matched cohort from the reimbursement database of the National Health Insurance for the study. The matched pairs of ever users and never users of metformin were derived from the unmatched cohort based on propensity scores (MBT: malignant brain tumors).
Figure 1. The step-by-step procedures followed to create an unmatched original cohort and a matched cohort from the reimbursement database of the National Health Insurance for the study. The matched pairs of ever users and never users of metformin were derived from the unmatched cohort based on propensity scores (MBT: malignant brain tumors).
Biomolecules 11 01226 g001
Table
Table 1. Characteristics of metformin never users and ever users and standardized difference between the two groups.
Table 1. Characteristics of metformin never users and ever users and standardized difference between the two groups.
VariablesUnmatched CohortMatched Cohort *
Never UsersEver UsersStandardized
Difference		Never UsersEver UsersStandardized
Difference
(n = 16,222)(n = 15,3429) (n = 16,222)(n = 16,222)
n%n% n%n%
Demographic and Basic Data
Age ** (years)63.62 ± 10.4361.84 ± 10.03−17.7563.62 ± 10.4363.83 ± 9.792.72
Sex (men)929857.3282,57553.82−7.72929857.32923856.95−0.98
Occupation
I633639.0659,85339.01 633639.06635939.20
II322919.9135,28623.008.08322919.91317719.58−0.81
III341021.0232,17520.970.07341021.02342321.100.36
IV324720.0226,11517.02−8.61324720.02326320.110.14
Living Region
Taipei545333.6148,38831.54 545333.61541433.37
Northern165810.2217,38611.333.74165810.22168610.390.59
Central284117.5128,06918.292.13284117.51286617.670.37
Southern280617.3026,17417.06−0.65280617.30285217.580.82
Kao-Ping and Eastern346421.3533,41221.781.22346421.35340420.98−0.75
Major Comorbidities
Hypertension (401–405)13,30982.04125,95582.090.2313,30982.0413,31582.080.33
Dyslipidemia
(272.0–272.4)		11,72372.27127,38783.0328.4511,72372.2711,75172.440.72
Obesity (278)4402.7169574.5310.004402.714112.53−1.13
Diabetes-related Complications
Nephropathy
(580–589)		566634.9342,45727.67−17.80566634.93555734.26−1.80
Eye diseases
(250.5, 362.0, 369,
366.41, and 365.44)		301118.5649,86132.5032.53301118.56285417.59−3.07
Stroke (430–438)540133.2945,89929.92−8.11540133.29535232.99−0.59
Ischemic Heart Disease (410–414)777347.9270,78946.14−3.78777347.92780048.080.48
Peripheral arterial disease (250.7, 785.4, 443.81 and 440–448)377723.2839,98226.066.61377723.28368822.73−1.42
Antidiabetic drugs
        Insulin13518.3335712.33−30.6113518.3311377.01−6.63
Sulfonylurea11,79072.68111,54672.705.6611,79072.6812,19975.205.89
Meglitinide13408.2660323.93−19.3413408.2613178.12−0.59
        Acarbose183511.3183975.47−20.71183511.31184111.35−1.13
Rosiglitazone4792.9575994.9510.834792.955093.140.53
Pioglitazone4012.4740492.64−20.714012.474292.64−1.13
Commonly encountered comorbidities
Chronic obstructive pulmonary disease (490–496)808749.8574,98748.87−2.40808749.85824650.832.11
Tobacco abuse
(305.1, 649.0 and 989.84)		4602.8461454.016.674602.844582.82−0.05
Alcohol-related diagnoses
(291, 303, 535.3, 571.0–571.3 and 980.0)		12857.9210,9737.15−4.2312857.9211917.34−2.39
Ocular pterygium
(372.40–372.44)		8975.5389905.861.448975.538945.510.02
Medications that are commonly used in diabetes patients
    Angiotensin converting enzyme inhibitor/angiotensin receptor blocker11,29869.65112,72073.478.8511,29869.6511,28069.54−0.15
Calcium channel blocker10,21562.9792,51860.30−5.6510,21562.9710,26563.280.79
    Statin876854.05101,37166.0726.41876854.05873053.82−0.33
    Fibrate554934.2166,52143.3620.08554934.21547433.74−0.81
    Aspirin933357.5395,05861.969.38933357.53929057.27−0.32
* The matched cohort was created from the unmatched cohort based on propensity score; ** age is denoted by mean ± standard deviation. Refer to Materials and Methods for the classification of occupation. Parentheses include the diagnostic codes of diseases according to the International Classification of Diseases, Ninth Revision, Clinical Modification.
Table
Table 2. Incidence rates of malignant brain tumors and hazard ratios by metformin exposure.
Table 2. Incidence rates of malignant brain tumors and hazard ratios by metformin exposure.
Cohort/Metformin UseIncident Cases of Malignant Brain TumorsCases FollowedPerson-YearsIncidence Rate (Per 100,000 Person-Years)Hazard Ratio95% Confidence Intervalp Value
Unmatched Cohort
Never Users2716,22272,755.3837.111.000
Ever Users155153,429724,547.5021.390.574(0.381–0.863)0.0077
Tertiles of cumulative duration of metformin therapy (months)
Never Users2716,22272,755.3837.111.000
<27.135950,605178,095.3133.130.897(0.567–1.421)0.6440
27.13–58.335850,628248,115.7123.380.623(0.395–0.984)0.0426
>58.333852,196298,336.4812.740.316(0.192–0.518)<0.0001
Matched Cohort
Never Users2716,22272,755.3837.111.000
Ever Users916,22276,004.8911.840.317(0.149–0.673)0.0028
Tertiles of cumulative duration of metformin therapy (months)
Never Users2716,22272,755.3837.111.000
<27.003534318,587.1316.140.427(0.129–1.412)0.1632
27.00–58.405536126,025.9719.210.509(0.196–1.322)0.1657
>58.401551831,391.793.190.087(0.012–0.639)0.0164
Table
Table 3. Sensitivity analyses.
Table 3. Sensitivity analyses.
Model/Metformin UseIncident Cases of Malignant Brain TumorsCases FollowedHazard Ratio95% Confidence Intervalp Value
I.  After excluding patients who had received any two consecutive metformin prescriptions spanning a period of four or more months
Never Users2716,2221.000
Ever Users4851,6160.568(0.355–0.911)0.0188
Tertiles of cumulative duration of metformin therapy (months)
Never Users2716,2221.000
<27.131716,8790.979(0.527–1.818)0.9463
27.13–58.331614,0140.675(0.363–1.254)0.2134
>58.331520,7230.324(0.172–0.611)0.0005
II.  After excluding patients who happened to be treated with incretins during follow-up
Never Users2715,2371.000
Ever Users151117,1710.703(0.467–1.059)0.0916
Tertiles of cumulative duration of metformin therapy (months)
Never Users2715,2371.000
<27.136042,6001.002(0.634–1.582)0.9940
27.13–58.335338,3280.710(0.446–1.128)0.1471
>58.333836,2430.428(0.261–0.703)0.0008
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Tseng, C.-H. Metformin and Risk of Malignant Brain Tumors in Patients with Type 2 Diabetes Mellitus. Biomolecules 2021, 11, 1226. https://doi.org/10.3390/biom11081226

AMA Style

Tseng C-H. Metformin and Risk of Malignant Brain Tumors in Patients with Type 2 Diabetes Mellitus. Biomolecules. 2021; 11(8):1226. https://doi.org/10.3390/biom11081226

Chicago/Turabian Style

Tseng, Chin-Hsiao. 2021. "Metformin and Risk of Malignant Brain Tumors in Patients with Type 2 Diabetes Mellitus" Biomolecules 11, no. 8: 1226. https://doi.org/10.3390/biom11081226

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop