Metformin: not only per os

KEYWORDS:

• Antidiabetic biguanides
• cancer
• diabetes
• metformin
• phenformin
• routes of administration

1. Introduction

‘Metformin is everywhere’[1] not only because of its current position as a most accessible antidiabetic biguanide. R. Romero et al [1] called metformin a 21st century aspirin reminding its spacious uses: apart from obvious role as antidiabetic in type 2 diabetes mellitus it may be applied for prediabetes control, polycystic ovarian disease treatment, prevention of neonatal macrosomia, metabolic disorders in obese children and adolescents, etc. In particular, there is a widely discussed and potentially fruitful area of oncology and gerontology, although here the effects still need to be proved in randomised trials [2–4]. Metformin may also be used as a part of preventive measures for several cardiovascular and cerebrovascular (including decrease in cognitive function) conditions and for some other diseases.

In the current Editorial, we speak about a different matter. We discuss whether we should limit ourselves to oral use of the drug, what is known of possible routes bypassing the gastrointestinal tract, and the data in favour of these routes. For an introductory reasoning, it is enough to draw just two arguments here. First of all, metformin influences many reactions and processes (proliferation, apoptosis, angiogenesis, oxidative stress etc.) in cell lines, that is in vitro. Therefore, the common explanations and mechanisms of its antidiabetic effect may not be the same when the drug is used with other intentions or via a different route of administration, although this concept certainly requires further investigation.

The second, even more important, argument for ‘bypassing the oral route’ lies in gradually accumulating specific data. The further part of the Editorial will describe what is already known and what may possibly be.

1.1. Available data

There are at least two ways to analyse the antidiabetic effect, which metformin exerts via central nervous system. The first one is represented by metformin infusion into duodenum (bypassing the portal vein system). In a rat diabetes mellitus model it leads to isolated activation of duodenal adenosine monophosphate kinase (AMPK) with subsequent suppression of glucose production via duodenal protein kinase A GLP-1 receptors/CNS centres/liver axis [5]. The second, more direct, way is metformin injection into lateral ventricles of the brain, which, apart from other effects (including an increase of sensitivity to insulin and less pronounced shift in glucose tolerance), may cause non-hypothalamic AMPK activation-mediated anorexia in experimental animals [6].

As another example of ‘non-traditional’ metformin route of insertion we would like to reference recent study, in which the topical use of metformin was aimed to heal skin lesions in 3 months-old C57Black/6 mice and skin regeneration/aging delay in 18 months-old Sprague Dawley rats [7]. Therefore, a potential sterile solution or gel-like metformin formula might be used (after proper clinical trials and all due procedures are finished) in dermatological practice including skin tumours [8] and diabetes mellitus patients. There are, however, certain doubts about the latter (in particular, considering the diabetic foot) as metformin’s antiproliferative effect may impair the regeneration of diabetic ulcer [9].

A prototype of mentioned gel-like formula with 0.5–1.5% of biguanide was studied in chronic periodontitis patients (in which inflammatory process have spread from dental pulp to the roots) due to ability of metformin to limit the loss of alveolar bone tissue. It turned out that the use of this formula for 3 to 6 months via gel injections into periodontal pockets was able to decrease the severity of intra-bone lesions and improve some other parameters affected by periodontitis. The researchers came to conclusion that local (subgingival) metformin application may have some advantages over systemic one in terms of treatment effectiveness, therapeutic concentration and adverse effects rate [10]. Another point revealed by this study was higher effectiveness of 1% metformin gel injection into periodontal pockets for periodontitis control compared to 0.5% and 1.5% formulas [10]. The lack of linear dose-dependent effect in local (subgingival) metformin application may point at some peculiar characteristics of this route needing further analysis.

1.2. Potential application areas

Of course, we could predict many other areas for metformin ‘non-standard’ use like e.g. its application in form of eye drops for a whole spectrum of eye conditions in diabetes mellitus and metabolic syndrome patients. Apart from retinopathy and maculopathy these patients may develop the open angle glaucoma. The latter incidence may decrease by 20% in patients receiving a daily peroral metformin dose of 2 g, while rather often this effect is accompanied by gastrointestinal problems [11].

Consequently, the metformin suppositories could be used to avoid gastrointestinal complications seen in patients taking tablet form of this medicine. Here, we should not underestimate also the potential benefits of topical metformin use in inflammatory gastrointestinal tract conditions, which was demonstrated by intrarectal biguanide infusions (50 mg/kg for 16 days) in an inflammatory bowel disease mouse model [12]. Even without additional examples we could suppose different, perhaps better, effects of topical metformin on diabetes mellitus patients intestinal microbiota (which is currently a subject of intense research) in comparison to more traditional routes.

A special (and considerably large) field of potential non-routine insertion of metformin lies in cancer area. Below we provide (in addition to the paper [8]) a short summary of currently available data. The results of in vitro experiments suppose the ability of a gel-like structure composed of metformin molecules associated with cellulose nanofibers slow down the migration (and therefore metastasising and dissemination) of melanoblastoma cells [13]. Taking into account the data on lower colorectal cancer incidence in metformin-treated diabetic patients, there is a possibility of using fatty- (supposedly a better variant) or polyethylene glycol (PEG)-based suppositories at least in patients with colorectal adenomas with or without diabetes mellitus as an alternative to standard drug form [14]. The gel-like injectable formula containing a mixture of metformin and chemotherapy (5-fluorouracil) was studied in colorectal cancer mouse model [15], while an intravenous liposomal combination of metformin and anthracycline (epirubicin) had been rather efficient in transplanted S180 mouse sarcoma model [16]. Moreover, the term ‘oncobiguanides’ (although hardly as yet justified) was proposed in contrast to ‘diabetobiguanides’ to distinguish drugs with potentially higher anticancer activity and, perhaps, toxicity, which may be inhaled (in lung cancer cases), applied to the skin (in melanoma), taken as suppositories (colorectal cancer) or, in some patients, injected directly into the tumour [17].

It cannot be excluded that metformin may also be administered into uterine cavity (like levonorgestrel-releasing intrauterine system Mirena or in some other way) as monotherapy or in combination with progestin to treat a simple or atypical endometrial hyperplasia, pre-cancer state for endometrial cancer, as there is already some data on clinical activity of oral biguanide. All these facts and examples should be a subject of thorough discussion and studies. It should be mentioned, though, that there is already some experimental data on intratumoral metformin injection [18] obtained independently from other sources [17], and, besides, successful results were received with metformin intravesical treatment of implanted bladder cancer in mouse model [19].

1.3. Conclusions

Of course, the whole issue of these alternative routes (see Table 1) is a place for considerable doubt as ‘the traditional one works quite well’. However, there is a large population of patients developing gastrointestinal adverse effects and cobalamin deficiency with possible progression to neuropathy even on slow-release oral form, metformin XR. This state of metformin intolerance in diabetes mellitus treatment requires sometimes the shift to another, metformin-free, therapy variants. Besides well-known ‘non-biguanide’ treatments, the use of phenformin, potentially more active (while, unfortunately, more toxic) drug, is often implicated as an example of substitute.

Table 1. Potential routes of metformin administration (Already studied or proposed).

Route	Present state of implication
Oral use	Routine clinical practice
Injections into lateral brain ventricles	Experimental study in animals [6]
Topical use in skin lesions	Experimental studies in animals [7–9]
Injections in periodontal pockets at chronical periodontitis	Clinical approbation in patients [10]
Rectal biguanide infusions in an inflammatory bowel disease	Mouse model [12]
Rectal suppositories	Bioavailability of metformin-containing formulation is evaluated in human volunteers [14]
Subcutaneous injections	Peritumoral injections of hydrogel containing metformin and 5-fluorouracil were tested in BALB/c mice inoculated with colon adenocarcinoma C26 cells [15]
Intravenous injections	Liposomal combination of metformin and anthracycline (epirubicin) is described in transplanted S180 mouse sarcoma model [16]
Intratumoral injections	Xenografts of prostate cancer and melanoma cells grown in female mice[18]
Intravesical administration of metformin solution	Preclinical murine model of bladder cancer [19]
Suggested but not yet studied approaches
Inhalation [17]
Eye drops^This paper
Intrauterine administration^This paper

Sublingual rout, at least as we are aware, was not so far mentioned

In the discussions which can coexist with the development of this (‘not per mouth’) topic, one also must take into account metformin pharmacokinetics, its different serum and tissue concentrations when used via different routes, the gender differences in its activity, and the fact that oral hypolipidemic and antidiabetic (including metformin) medications non-adherence is usually associated with flawed compliance with hormonal therapy of cancer [20], which should not be always the case when the drug is inserted via unconventional ways.

In summary, we may state as a final conclusion that there are several examples of non-oral metformin experimental use in different conditions and models. Along with hormone-associated (endocrine) pathologies treatment, this approach may find its implementation, if continued, in oncology, ophthalmology, dermatology and some other areas of medicine. Generally, the abovementioned data may be useful for anyone interested in endocrinology and diabetology, while willing also to expand their borders of knowledge beyond their favourite topics.

Declaration of interest

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose

Additional information

Funding

This manuscript was not funded.