Inflammatory pain (ie, pain following trauma, surgery, or accompanying rheumatic/arthritic diseases) represents the most commonly treated clinical pain state. Depending on the intensity of pain, it can be managed with different conventional analgesic drugs (nonsteroidal antiinflammatory drugs [NSAIDs] or opioid analgesics), whereas in chronic pain states, adjuvant analgesics (eg, antiepileptic drugs) can be used alongside conventional analgesic drugs.1,2
Recently, it was demonstrated that metformin, a widely used first-line antihyperglycemic agent for type 2 diabetes therapy,3 possesses analgesic properties in the models of inflammatory and neuropathic pain.4–7 In addition, a couple of studies have suggested that metformin could be potentially useful in the treatment of fibromyalgia and lumbar radiculopathy pain in humans.8,9 Metformin is an activator of the adenosine monophosphate-activated kinase (AMPK), which was shown to play a role in modulating nociception.7,10 Epidemiological surveys have indicated that arthritic inflammatory disorders are highly prevalent among diabetic patients and that more than half of patients with diabetes mellitus have some form of arthritis.11 Among them, certain painful arthritic/rheumatic disorders represent relatively specific complications of diabetes mellitus, whereas other forms of arthritis (including osteoarthritis and possibly rheumatoid arthritis) tend to be more common or more severe in diabetic patients.12–14 Thus, the coadministration of metformin and analgesic drugs is likely to be prevalent among diabetic patients. Considering that metformin has displayed efficacy in different pain models, the investigation of the manner in which metformin and analgesics interact, with regard to pain relief, could provide some useful insights into the management of pain in diabetic patients.
In this study, we aimed to examine the effects of 2-drug combinations of metformin with ibuprofen and aspirin (widely used NSAIDs), tramadol (an opioid analgesic), and pregabalin (an adjuvant analgesic drug) in a rat model of somatic inflammatory hyperalgesia and to determine the type of interaction between the components.
This manuscript adheres to the ARRIVE guidelines. All experiments were approved by the Institutional Animal Care and Use Committee of the Faculty of Pharmacy, University of Belgrade and were carried out in compliance with the guidelines of European Communities Council Directive 2010/63/EU on the use of animals for scientific purposes. The experiments were performed on male Wistar rats (weighing 180–220 g), obtained from the Military Academy Breeding Farm, Belgrade, Serbia. The animals were housed in groups of 4 in home cages and maintained on a 12/12-h light/dark cycle at 22 ± 1°C and 60% relative humidity. Food and water were freely available, except during the experimental procedure. The experiments were conducted in a blinded manner, between 8:00 am and 4:00 pm to avoid diurnal variations in behavioral tests. A total number of 310 rats were used in the study.
Drugs and Their Administration
Metformin hydrochloride (a gift from Galenika AD Beograd, Belgrade, Serbia) and tramadol hydrochloride (Trodon, solution for injection, Hemofarm AD, Vršac, Serbia) were dissolved/diluted in isotonic saline and applied intraperitoneally (i.p.) in a volume of 2 mL/kg body weight. Ibuprofen (Pharmagen GmbH, Frankfurt, Germany), aspirin (Aspirin, Bayer Bitterfeld GmbH, Bitterfeld-Wolfen, Germany), and pregabalin (Lyrica, Pfizer Manufacturing Deutschland GmbH, Betriebsstätte Freiburg, Germany) were suspended in distilled water and administered by oral gavage in a volume of 2 mL/kg body weight.
Carrageenan λ (Sigma-Aldrich Chemie GmbH, Munich, Germany) was dispersed in saline (1% m/V) and applied intraplantarly to rats in a volume of 100 μL by using a 1-mL syringe and 26-gauge needle.
Model of Localized Inflammation
Inflammation of the rat hind paw was induced by an intraplantar injection of carrageenan. Carrageenan-induced inflammation is followed by the development of nociceptive hypersensitivity (hyperalgesia) and edema15–17 and was used to test the antihyperalgesic effects of drugs/drug combinations and the antiedematous effect of metformin.
Assessment of Antihyperalgesic Effects.
The development of mechanical hyperalgesia following carrageenan injection and the antihyperalgesic effects of the examined drugs and 2-drug combinations were assessed by measuring the paw withdrawal thresholds (PWTs) using an electronic Von Frey anesthesiometer (IITC Life Science, Woodland Hills, CA).17 Rats were placed in transparent, plastic boxes on top of a metal grid and allowed to acclimatize for 30 minutes before testing. The mechanical stimulus was delivered using a plastic, semiflexible filament coupled with a force transducer. The tip of the filament was applied perpendicularly to the plantar surface of the right hind paw, and the pressure was gradually increased. The force (in grams) required to elicit brisk paw withdrawal was automatically recorded on a digital screen. The average of 4 consecutive PWT measurements was used for further calculations.
Pretreatment PWTs were measured before the carrageenan injection and drug/drug combination administration. Posttreatment PWTs were measured 60, 90, 120, 150, 180, 240, and 300 min after the carrageenan injection. Pre- and posttreatment PWTs were always measured on the same (right) hind paw. Metformin, ibuprofen, and pregabalin were administered at the same time as carrageenan. Aspirin and tramadol were applied after the induction of inflammation (30 and 45 min after the carrageenan injection, respectively). The time points when the drugs were administered were chosen according to the data about the time course of carrageenan-induced hyperalgesia and the time course of antihyperalgesic effects produced by the examined drugs.7,18–21
The results are expressed as the difference (df) between pre- and posttreatment PWTs according to the following formula:17
The percent of antihyperalgesic (%AH) activity produced by a drug/drug combination was calculated according to the following formula:
The ED50 values (doses expected to produce 50% AH) with 95% confidence intervals were calculated from corresponding log dose-response curves.22,23
Assessment of Antiedematous Effects.
The antiedematous activity of metformin was assessed with a plethysmometer (Ugo Basile, Comerio, Italy), by measuring the increase in paw volume following the carrageenan injection, as described previously.16 Pretreatment paw volumes were measured before the carrageenan injection and metformin administration. The posttreatment paw volumes were measured 60, 90, 120, 150, 180, 240, and 300 min after the induction of inflammation. The results are expressed as the difference (dV) between post- and pretreatment paw volumes according to the following formula:
The measurements were repeated 2 times at each time point and the average dV of each rat was used for further calculations.
The percent of antiedematous activity (%AE) was calculated according to the following formula:16
Analysis of the Interactions Between Metformin and Analgesics in the Carrageenan-Induced Hyperalgesia Model
The interactions between metformin and analgesics were evaluated by isobolographic analysis at the ED50 level of effect, as described previously.21,24 In brief, the ED50 value of each drug was obtained from the corresponding log dose-response curve. In combination experiments, metformin and ibuprofen/aspirin/tramadol/pregabalin were coadministered in fixed-dose fractions of their ED50. For drug combinations, the experimental ED50 (ED50 mix) was determined by the linear regression analysis of the log dose-response curve and compared with a theoretical additive ED50 (ED50 add). When the drug combination produced an ED50 mix that was significantly lower than the ED50 add, it was interpreted as a supraadditive (synergistic) interaction between the drugs.24
In addition, an interaction index (γ) was used to describe the magnitude of interaction.21,25
Analysis of the Duration of Antihyperalgesic Effects of Drugs and Drug Combinations
To compare the duration of the antihyperalgesic effect of individually applied drugs with the duration of the effect produced by the same drug applied in combination, the slopes of the %AH-ΔAUC regression lines were calculated as previously described.21 In brief, the areas under the time-df curves (AUC) were calculated for the control (AUCC) and the drug/drug combination groups (AUCD), and the differences were calculated as ΔAUC = AUCC − AUCD.23 The ΔAUC for each drug/drug combination group is presented as a function of its peak %AH (ΔAUC = slope × %AH + intercept). In the %AH-ΔAUC regression line, the slope is the relative measure of the duration of the drug/drug combination effect; the treatment with a significantly greater slope exerts an effect of longer duration than that with a smaller slope.21,26
In addition, a high correlation coefficient of the %AH-ΔAUC regression line indicates that the duration of the effect of the drug/drug combination treatment is dose dependent.21,26
The rotarod test was used to evaluate the effects of metformin, tramadol, and pregabalin, as well as the effects of the metformin-tramadol and metformin-pregabalin combinations on motor coordination or sedation.27 The test was performed using a rotarod apparatus (Treadmill for rats 47700; Ugo Basile, Milano, Italy), rotating at a constant speed of 15 rpm. The animals were trained to drive the rotarod for 2 days. On the day of the experiment, only those rats that could remain on the rod for 180 s on 2 consecutive trials were used. The posttreatment latency to remain on the rotating rod was recorded at 7 time points, which correspond to the time points when the antihyperalgesic effects of drugs/drug combinations were measured in the carrageenan-induced hyperalgesia model. The cutoff time was 180 s.
The statistical analysis was performed using SigmaPlot 11 (Systat Software Inc, Richmond, CA) and SPSS 18 for Windows (IBM SPSS Statistics, Chicago, IL). The results are presented as mean group values ± SEM (confidence intervals, CIs).
The data from the carrageenan-induced hyperalgesia and edema model were assessed for normality (Shapiro-Wilk test) and equality of variance (Mauchly test of sphericity and Levene test) and were analyzed by a 2-way repeated-measures analysis of variance (ANOVA) followed by the Tukey Honest Significant Difference (HSD) test for between-group comparisons. We examined the effects of the type of treatment (as the between-subject factor) and time after treatment administration (as the within-subject factor), as well as the interaction between these 2 factors. The statistical assumptions of normality and equality of variance were not violated in most treatment groups. The only exceptions were in the analysis of the antihyperalgesic effects of pregabalin and metformin-tramadol combination, and in the analysis of the antiedematous effects of metformin, where the assumption of sphericity was violated (all P ≤ .001) and the Greenhouse-Geisser correction was applied to determine the statistical significance of the within-subject factor (time after treatment administration). In addition, we detected a significant interaction (all P ≤ .005) between the 2 tested factors in 3 cases, in the analysis of the antihyperalgesic effects of pregabalin and metformin-pregabalin combination, and in the analysis of the antiedematous effects of metformin. In these cases, we performed an additional 1-way ANOVA, followed by Tukey HSD test, for every time point to determine which drug/drug combination–treated group was significantly different from the control group.
The data from the rotarod test were not normally distributed (Shapiro-Wilk test) and were analyzed with the exact Mann-Whitney U test.
All pharmacological computations were done according to Tallarida et al22–24 using computer programs Pharm PCS (Micro-Computer Specialists, Philadelphia, PA) and Pharm Tools Pro (The McCary Group, Schnecksville, PA). The difference between theoretical ED50 (ED50 add) and experimental ED50 (ED50 mix) for drug combinations was examined by a modified t test.24 The slopes of the %AH-ΔAUC regression lines were compared using the test for parallelism (which utilizes the Student t test to compare the slopes of 2 regression lines).23 A P value of <.05 was considered statistically significant, except in the test for parallelism where the Bonferroni correction was used to adjust for multiple comparisons. After the application of the Bonferroni correction, a P value of <.013 was considered statistically significant when comparing the duration of effects of metformin with the duration of effects of other tested analgesics, whereas a P value of <.025 was considered significant when comparing the duration of effects of individual drugs with the duration of effects of a drug combination. The first corrected P value was obtained by lowering the critical significance probability to P < .05/4, where 4 represents the number of multiple comparisons made in the analysis of the duration of effects of individual drugs (the duration of metformin’s effects compared with the duration of effects of the 4 tested analgesics). The second corrected P value was obtained by lowering the critical significance probability to P < .05/2, for 2 comparisons that were made in the analysis of the duration of effects of drug combinations. For a given 2-drug combination (consisting of drug A and drug B), the following comparisons were made: duration of effects of drug A versus duration of effects of drug combination, and duration of effects of drug B versus duration of effects of drug combination.
The size of the experimental groups in the carrageenan-induced hyperalgesia/edema model was 11 rats for the control group, 6 rats for the drug/drug combination–treated groups in the hyperalgesia model, and 11 rats for the metformin-treated groups in the edema model. In the rotarod test, the size of the experimental groups was 7 rats for the control and 6–7 rats for the drug/drug combination–treated groups. The group sizes were chosen on the basis of our previous experience with these tests and are consistent with literature data. In the carrageenan-induced hyperalgesia model, a sample size of 6 animals per test group provided adequate statistical power (1−β ≥ 0.87) to detect a significant antihyperalgesic effect of 20% in every time point (effect size ≥ 0.61; α <.05). In the carrageenan-induced edema model, a larger sample size of 11 animals per test group was needed to obtain appropriate statistical power (1−β ≥ 0.81) to detect a significant antiedematous effect of 20% in later time points (150–240 min after induction of inflammation), when the maximal size of edema is usually achieved and which are more relevant for the assessment of antiedematous activity (effect size ≥ 0.48; α < .05). Antihyperalgesic/antiedematous effects in the range of 20% to 80% are needed for accurate determination of ED50 values (by linear regression analysis of log dose-response curves) and subsequent isobolographic analysis.22,23 Effect sizes were calculated for every time point using the control group average values for df/dV and their standard deviations. In addition, a sample size of minimum 6 animals per test group provided appropriate statistical power (1−β = 0.87) to detect a 20% decrease in the time spent on the rotarod apparatus (effect size = 1.7; α < .05). The effect size for the rotarod test was calculated using a standard deviation of 0 for the control group and a standard deviation of 30 for the treatment group (an effect of 20% reduction of the time spent on the rotarod apparatus and the values of the control/treatment group standard deviations were chosen on the basis of our previous experiments with drug treatments with known effects on motor coordination). Statistical power and effect sizes were calculated using G*Power 220.127.116.11 (University of Düsseldorf, Germany).
The Antihyperalgesic Effects of Metformin and Analgesics in a Model of Carrageenan-Induced Inflammatory Hyperalgesia
Metformin (50–200 mg/kg) produced a significant and dose-dependent antihyperalgesic effect in the model of inflammatory hyperalgesia induced by carrageenan (Figure 1A). The 2-way repeated-measures ANOVA revealed a significant effect of both type of treatment and time after metformin application on the df value determined using the electronic Von Frey apparatus (P < .001 for both factors; the P value for the interaction between 2 factors was .059). The Tukey post hoc P values were ≤.024 in all time points for most tested doses of metformin (Figure 1A). Most tested doses produced a maximal effect 90 min after the induction inflammation and the df values in metformin-treated groups at this time point were from 8.5 ± 1.3 g (95% CI: 5.9–11 g) to 18 ± 1.1 g (95% CI: 15–20 g), whereas the df value determined in the control group was 22 ± 1 g (95% CI: 20–24 g). The antihyperalgesic effects produced by metformin, at the 90-min time point, were from 21 ± 5% (95% CI: 11%–30%) to 62 ± 6% (95% CI: 50%–74%) (Figure 2) and the ED50 ± SEM value was 125 ± 7.1 mg/kg (Table 1).
Ibuprofen (25–150 mg/kg), aspirin (100–400 mg/kg), tramadol (0.5–5 mg/kg), and pregabalin (2.5–20 mg/kg) significantly reduced the mechanical hyperalgesia induced by carrageenan in a dose-dependent manner (Figure 1, B–E). A significant effect of the type of treatment and time after drug administration on the df value determined using the electronic Von Frey apparatus (P < .001 for both factors for all 4 tested analgesics) was detected using a 2-way repeated-measures ANOVA. The interaction between the 2 tested factors was not significant for the ibuprofen-, aspirin-, and tramadol-treated groups (all P ≥ .169) and the Tukey post hoc P values were ≤.042 in all time points for most tested doses of these analgesics (Figure 1, B–D). In the pregabalin-treated groups, we detected a significant interaction between the 2 tested factors (P = .002), so we performed an additional 1-way ANOVA, followed by Tukey HSD test, for every time point to determine which pregabalin-treated group was significantly different from the control group. One-way ANOVA detected a significant antihyperalgesic effect of pregabalin in every time point (all P < .001), and the Tukey post hoc P values were ≤ .024 in all time points for most tested doses of pregabalin (Figure 1E).
For all analgesics, maximal effects were achieved 90 min after the induction of inflammation with carrageenan. At the time of peak effects, the df values in groups that received the highest tested doses of analgesics were 5.9 ± 0.9 g (95% CI: 4.2–7.6 g) for ibuprofen, 8.5 ± 0.9 g (95% CI: 6.7–10 g) for aspirin, 6.9 ± 1.3 g (95% CI: 4.4–9.5 g) for tramadol, and 9.9 ± 0.9 g (95% CI: 8.1–12 g) for pregabalin, whereas the df value determined in the control group was 22 ± 1 g (95% CI: 20–24 g). The maximal antihyperalgesic effects of the highest tested doses of analgesics were 73 ± 4% (95% CI: 66%–81%), 62 ± 4.2% (95% CI: 53%–70%), 69 ± 5.9% (95% CI: 57%–80%), and 56 ± 3.9% (95% CI: 48%–63%) for ibuprofen, aspirin, tramadol, and pregabalin, respectively (Figure 2). The ED50 values of the examined analgesics were calculated from the corresponding log dose-response curves (Figure 2) and are presented in Table 1.
The Effects of Metformin on Carrageenan-Induced Edema
The 2-way repeated-measures ANOVA detected a nonsignificant effect of the type of treatment (metformin dose; P = .430) and a significant influence of time after metformin application (P < .001) on the dV value determined in the model of carrageenan-induced edema. In addition, there was a significant interaction between the 2 tested factors (P = .005), so we performed 1-way ANOVA, followed by Tukey HSD test, for every time point to determine the main effect of the type of treatment on carrageenan-induced edema. One-way ANOVA detected a significant antiedematous effect of metformin in 1 time point, 240 min after the induction of inflammation (P = .039). The post hoc analysis revealed that only the highest tested dose of metformin (200 mg/kg) produced a significant reduction of edema (the Tukey post hoc P value was .020; Figure 3). The antiedematous effect produced by this dose at the 240-min time point was 27 ± 5.4% (95% CI: 17%–38%).
Interactions Between Metformin and Analgesics in a Model of Somatic Inflammatory Hyperalgesia
Two-drug combinations of metformin and analgesics (ibuprofen, aspirin, tramadol, or pregabalin) administered in fixed-dose fractions of the ED50 (1/16, 1/8, 1/4, and 1/2) of each drug, produced significant and dose-dependent antihyperalgesic effects (Figure 4, A–D). The 2-way repeated-measures ANOVA revealed a significant effect of both type of treatment and time after drug combination application on the df value determined using the electronic Von Frey apparatus (P < .001 for both factors for all 4 tested combinations). The interaction between the 2 tested factors was not significant for the metformin-ibuprofen, metformin-aspirin, and metformin-tramadol treated groups (all P ≥ .233), and the Tukey post hoc P values were ≤.027 in all time points for most tested doses of metformin-analgesic combinations (Figure 4, A–C). In the metformin-pregabalin–treated groups, we detected a significant interaction between the 2 tested factors (P < .001), so we performed an additional 1-way ANOVA, followed by Tukey HSD test, for every time point to determine which metformin-pregabalin–treated group was significantly different from the control group. One-way ANOVA detected a significant antihyperalgesic effect of the metformin-pregabalin combination in every time point (all P < .001), and the Tukey post hoc P values were ≤.027 in all time points for most tested doses of the metformin-pregabalin combination (Figure 4D).
At the time of peak effects (90 or 120 min after the carrageenan injection), the df values in groups that received the highest tested doses of metformin-analgesic combinations were 9 ± 0.7 g (95% CI: 7.7–10 g) for the metformin-ibuprofen combination (90 min postcarrageenan), 8.1 ± 0.8 g (95% CI: 6.3–9.6 g) for the metformin-aspirin combination (90 min postcarrageenan), 9.5 ± 1 g (95% CI: 7.6–11 g) for the metformin-tramadol combination (90 min postcarrageenan), and 10 ± 0.8 g (95% CI: 8.5–12 g) for the metformin-pregabalin combination (120 min postcarrageenan). The df values determined in the control group were 22 ± 1 g (95% CI: 20–24 g) at 90 min and 26 ± 1.6 g (95% CI: 23–29 g) at 120 min after the carrageenan injection.
The ED50 mix values, calculated from the corresponding log dose-response curves, are presented in Table 1. For all the examined combinations, the ED50 mix values were significantly lower than the ED50 add values (P < .05 by modified t test), and the interaction index was smaller than 1, which indicates a synergistic interaction (Figure 5, A–D and Table 1). According to the values of the interaction index (Table 1), in all examined combinations, there was a similar degree of potentiation with approximately 5-fold reduction of doses of both drugs when the drugs were applied in combination compared with the doses that produced the same level of effect after individual administration. In addition, a graphic illustration on the isobolograms (Figure 5, A–D) shows that the 95% CIs of ED50 mix and ED50 add values do not overlap, which is also a confirmation of synergism between the tested drugs.
Analysis of the Duration of Effects of Metformin, Analgesics, and Metformin-Analgesic Combinations in a Model of Somatic Inflammatory Hyperalgesia
The duration of antihyperalgesic effects of metformin, analgesics, and metformin-analgesic combinations was expressed as the slopes of the %AH-ΔAUC regression lines (see Methods). The calculated slopes are presented in Table 2.
Based on the slope values of the %AH-ΔAUC regression lines of individually applied drugs, metformin exhibited a significantly longer antihyperalgesic effect than pregabalin (P = .011 by test for parallelism; Table 2), and a similar duration of antihyperalgesic effects as ibuprofen, aspirin, and tramadol (all P ≥ .024 by test for parallelism; Table 2).
According to the values of the corresponding slopes, the duration of metformin’s effects was unchanged when it was combined with analgesics (all P ≥ .030 by test for parallelism; Table 2). Most tested analgesics also demonstrated an unchanged duration of antihyperalgesic effects when they were applied in combination with metformin (all P ≥ .083 by test for parallelism; Table 2) with the exception of pregabalin, whose effects were longer in duration when it was combined with metformin (P = .008 by test for parallelism; Table 2).
In most tested combinations, the peak antihyperalgesic effects were achieved at the same time as the peak of individually applied metformin or analgesics (90 min after the induction of inflammation with carrageenan). The only exception was the metformin-pregabalin combination, where the maximal antihyperalgesic effects were achieved 120 min after the induction of inflammation (ie, 30 min later than when the components were applied individually).
Effects of Metformin, Tramadol, and Pregabalin on Rotarod Performance
The highest doses of metformin (200 mg/kg), tramadol (5 mg/kg), and pregabalin (20 mg/kg) that were used in the model of somatic inflammatory hyperalgesia had no significant effect in the rotarod test in rats (all exact P ≥ .628 in every time point by Mann-Whitney U test; Figure 6). In addition, the highest tested doses (consisting of 1/2 of the ED50 of individual drugs determined in the carrageenan-induced hyperalgesia model) of the metformin-tramadol and metformin-pregabalin combinations did not produce a significant effect in rotarod test (all exact P ≥ .628 in every time point by Mann-Whitney U test; Figure 6).
In this study, metformin (50–200 mg/kg; i.p.) produced a significant and dose-dependent reduction of carrageenan-induced inflammatory hyperalgesia. The highest tested dose of metformin had no significant influence on rotarod performance, indicating that the observed antihyperalgesic effects were not because of motor impairment of rats. Metformin showed comparable efficacy as the other tested analgesics in reducing inflammatory hyperalgesia and was more potent than aspirin, but less potent than ibuprofen, tramadol, and pregabalin. Our results are consistent with the recent findings of Russe et al7 who demonstrated that metformin (100 mg/kg; i.p.) is capable of reducing zymosan-induced inflammatory mechanical hyperalgesia and nociceptive behavior in the second (inflammatory) phase of the formalin test in mice. We extend the existing findings by providing evidence that metformin’s effects on inflammatory hyperalgesia are present over a wider range of doses and that they are dose dependent.
Contrary to its antihyperalgesic effects, in our study, metformin demonstrated considerably lower efficacy in reducing carrageenan-induced edema (only the highest tested dose of 200 mg/kg produced a slight reduction of edema). The present finding is somewhat surprising considering that other activators of the AMPK, like AICAR, have been shown to reduce inflammatory edema.7 In addition, metformin has demonstrated antiinflammatory properties in different models of inflammation (reviewed by Saisho in 2015),28 including the ability to inhibit the production of proinflammatory mediators (such as prostaglandins and certain cytokines like tumor necrosis factor-α)29 that are important for the generation of carrageenan-induced edema.15,30 However, it should be noted that metformin’s antiinflammatory effects (including the effect on proinflammatory mediator synthesis29) are usually observed with relatively high concentrations (in the millimolar range), so it is possible that the doses we used in our study do not achieve sufficient local concentrations in the paw (following i.p. administration) to reduce edema. Further study is needed to explore this possibility.
It is widely recognized that metformin acts as an activator of AMPK, an ubiquitous kinase involved in multiple cell functions. In neural cells of pain transmission/modulation system, it serves as a negative regulator of mitogen-activated protein kinases (MAPKs) and mammalian target of rapamycin (mTOR) kinase signaling pathways, which are known to play an important role in pain plasticity, that is, in the development of pain hypersensitivity.10,31,32 Activation of AMPK has been shown to be responsible for the antihyperalgesic effects of metformin, as well as for the antinociceptive/antiallodynic effects of other AMPK-activating agents (AICAR, resveratrol) in inflammatory pain models.7,33 These effects were further linked to the inhibition of MAPKs and mTOR kinase signaling pathways. AICAR reduced nociception in the second phase of the formalin test following systemic administration, by inhibiting different MAPKs in spinal cord neurons,7 whereas peripheral administration of resveratrol attenuated incision-induced mechanical allodynia through the inhibition of MAPK and mTOR kinase activity in primary sensory neurons.33 The inhibition of MAPKs and mTOR kinase could also be relevant for metformin’s ability to reduce carrageenan-induced hyperalgesia, because elevated MAPK and mTOR kinase activity in nociceptive pathways contributes to the development of hyperalgesia in this model34–37 and specific inhibitors of different MAPKs and the mTOR kinase are capable of attenuating carrageenan-induced hyperalgesia.34,38,39 Because metformin readily passes the blood-brain barrier,40 it is possible that (after systemic administration) it exerts its antihyperalgesic effects by activating AMPK at both the central and peripheral level of nociceptive pathways.
The main finding of this study is that metformin interacted synergistically with several conventional and adjuvant analgesics (ibuprofen/aspirin/tramadol/pregabalin) in reducing inflammatory hyperalgesia. There was a similar, approximately 5-fold, reduction of doses in all tested combinations. To the best of our knowledge, the interaction between metformin and tramadol or pregabalin in pain models has not been studied until now. However, the interactions between systemically administered metformin and NSAIDs were previously examined by Ortiz41 who demonstrated that metformin (30–180 mg/kg) antagonizes the antinociceptive effects of diclofenac during the second (inflammatory) phase of the formalin test in rats. The discrepancy between our findings and the results of the cited study could be attributed to the use of different inflammatory pain models or different NSAIDs. To this extent, the same study demonstrated that metformin has no significant influence on the antinociceptive effect produced by another NSAID (indomethacin) in the formalin test.41 Our findings might suggest that ibuprofen and aspirin are favorable NSAIDs for treating inflammatory pain in patients that are concomitantly using metformin.
The interaction between 2 drugs may be a consequence of different mechanisms and/or levels of drug action (pharmacodynamic interaction) or because of alterations of the pharmacokinetic properties of individual drugs (pharmacokinetic interaction). A pharmacodynamic interaction between metformin and the examined analgesics seems to be a plausible cause for the synergism observed in our study, considering that metformin possesses a distinct mechanism of antihyperalgesic action (activation of AMPK) compared with the other tested analgesics.
Ibuprofen and aspirin are nonselective cyclooxygenase-1 and -2 inhibitors and reduce inflammatory hyperalgesia by inhibiting prostaglandin synthesis. Carrageenan-induced inflammation is associated with increased prostaglandin production at the site of inflammation and spinal level, which jointly contribute to the development of hyperalgesia.15,42,43 So, both peripheral and central cyclooxygenase inhibition could contribute to the antihyperalgesic effects of ibuprofen and aspirin.
Tramadol’s mechanism of antinociceptive action includes an opioid (μ-opioid receptor [MOR] activation) and nonopioid component (inhibition of serotonin/noradrenaline reuptake).44 Activation of MORs at central and peripheral sites has been shown to reduce carrageenan-induced hyperalgesia.45–47 In addition, tramadol’s ability to inhibit serotonin/noradrenaline reuptake could further increase the antihyperalgesic effects produced by MOR activation, seeing as putative serotonin/noradrenaline reuptake inhibitors are known to possess antihyperalgesic properties in the model of carrageenan-induced inflammation.48
Pregabalin is a specific ligand/modulator of the α2-δ auxiliary subunit of voltage-gated calcium channels and reduces nociception by inhibiting the release of certain nociceptive mediators.49 In models involving peripheral inflammation, pregabalin decreased the spinal release of glutamate and sensory neuropeptides (substance P and calcitonin gene–related peptide).50,51 These spinal mediators also contribute to the development of carrageenan-induced hyperalgesia,52,53 so the inhibition of their release could underline pregabalin’s effects in this model.
As discussed above, a synergistic interaction between 2 drugs could also be a consequence of pharmacokinetic interactions. We did not examine the pharmacokinetic interactions, so they cannot be excluded. However, metformin has a low potential for pharmacokinetic interactions, because it is neither significantly metabolized nor bound to plasma proteins.54 We observed that the duration of metformin’s antihyperalgesic effects in 2-drug combinations was not significantly different from the duration of effects of individually applied metformin. The duration of ibuprofen’s, aspirin’s, and tramadol’s antihyperalgesic effects was also unchanged when they were combined with metformin. If a pharmacokinetic interaction caused enhancements of the pharmacological effect, a prolongation of the effects of the drug combinations would be expected. On the other hand, we observed that pregabalin’s effects were longer in duration when it was combined with metformin and that the peak antihyperalgesic effects were achieved at a later time point than after individual drug administration. So, a pharmacokinetic interaction could be an additional mechanism contributing to the potentiation of antihyperalgesic effects in the metformin-pregabalin combination.
The preclinical examination of side effects of metformin-analgesic combinations may have less predictive value than the examination of their analgesic effects. However, we demonstrated that the highest tested doses of the metformin-tramadol and metformin-pregabalin combinations had no significant effect in the rotarod test, indicating that there is no enhancement of their effects on sedation/motor coordination of rats. An increased occurrence of side effects with these combinations in humans also seems unlikely because (1) the doses of components are significantly reduced and (2) the components have mostly different side-effect profiles, so the potentiation of their individual adverse effects seems less possible.3,44,49,55 Certain drug-specific side effects could also be attenuated (eg, metformin’s ability to decrease appetite/body weight might reduce pregabalin-induced weight gain). For all these reasons, the examined combinations could have a better safety profile than the corresponding drugs used in monotherapy.
The doses of metformin that were used in this study (both individually [50–200 mg/kg] and in combination with analgesics [8–63 mg/kg]) correspond to a human dose range of approximately 80–2000 mg56 and encompass metformin doses that are therapeutically used for treating type 2 diabetes mellitus. This may suggest that metformin-analgesic combinations could be used for the simultaneous treatment of diabetes mellitus and inflammatory pain and that lower doses of the examined analgesics might be sufficient for achieving adequate pain relief.
The reduction of aspirin doses observed in the metformin-aspirin combination could have some additional benefits. Low aspirin doses are widely used for their cardioprotective effects by patients with increased cardiovascular risk (eg, diabetic patients). The doses of aspirin that were used in combination with metformin in our study (15–122 mg/kg) correspond to human doses of approximately 150–1200 mg56 and correlate well with aspirin doses that have demonstrated cardioprotective effects in clinical trials.57 Therefore, a metformin-aspirin combination could be particularly suitable for patients suffering from type 2 diabetes mellitus, because it may enable achievement of multiple therapeutic goals (glucoregulation, pain relief, and cardioprotection) in this population with fewer drugs.
In conclusion, we demonstrated that metformin synergizes with ibuprofen/aspirin/tramadol/pregabalin to reduce hyperalgesia in a somatic inflammatory pain model. Our findings may suggest that in patients who are already receiving metformin therapy, lower (and potentially safer) doses of ibuprofen/aspirin/tramadol/pregabalin might be sufficient for achieving satisfactory pain relief. Metformin-aspirin combination might be particularly useful because it may achieve multiple therapeutic goals (glucoregulation, pain relief, and cardioprotection).
The authors thank Galenika AD Beograd (Belgrade, Serbia) for their kind donation of metformin hydrochloride.
Name: Uroš B. Pecikoza, BPharm.
Contribution: This author helped conduct the study, analyze the data, and write the manuscript.
Name: Maja A. Tomić, PhD.
Contribution: This author helped design the study and analyze the data.
Name: Ana M. Micov, PhD.
Contribution: This author helped conduct the study and analyze the data.
Name: Radica M. Stepanović-Petrović, PhD.
Contribution: This author helped design the study and analyze the data.
This manuscript was handled by: Jianren Mao, MD, PhD.
1. Gilron I, Jensen TS, Dickenson AH. Combination pharmacotherapy for management of chronic pain: from bench to bedside. Lancet Neurol. 2013;12:1084–1095.
2. Woolf CJ: American College of Physicians; American Physiological Society. Pain: moving from symptom control toward mechanism-specific pharmacologic management. Ann Intern Med. 2004;140:441–451.
3. Inzucchi SE, Bergenstal RM, Buse JB, et al. Management of hyperglycemia in type 2 diabetes, 2015: a patient-centered approach. Update to a position statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetologia. 2015;58:429–442.
4. Ma J, Yu H, Liu J, Chen Y, Wang Q, Xiang L. Metformin attenuates hyperalgesia and allodynia in rats with painful diabetic neuropathy induced by streptozotocin. Eur J Pharmacol. 2015;764:599–606.
5. Mao-Ying QL, Kavelaars A, Krukowski K, et al. The anti-diabetic drug metformin protects against chemotherapy-induced peripheral neuropathy in a mouse model. PLoS One. 2014;9:e100701. doi:10.1371/journal.pone.0100701.
6. Melemedjian OK, Asiedu MN, Tillu DV, et al. Targeting adenosine monophosphate-activated protein kinase (AMPK) in preclinical models reveals a potential mechanism for the treatment of neuropathic pain. Mol Pain. 2011;7:70. doi:10.1186/1744-8069-7-70.
7. Russe OQ, Möser CV, Kynast KL, et al. Activation of the AMP-activated protein kinase reduces inflammatory nociception. J Pain. 2013;14:1330–1340.
8. Bullón P, Alcocer-Gómez E, Carrión AM, et al. AMPK phosphorylation modulates pain by activation of NLRP3 inflammasome. Antioxid Redox Signal. 2016;24:157–170.
9. Taylor A, Westveld AH, Szkudlinska M, et al. The use of metformin is associated with decreased lumbar radiculopathy pain. J Pain Res. 2013;6:755–763.
10. Price TJ, Dussor G. AMPK: An emerging target for modification of injury-induced pain plasticity. Neurosci Lett. 2013;557(Pt A):9–18.
11. Centers for Disease Control and Prevention. Arthritis as a potential barrier to physical activity among adults with diabetes—United States, 2005 and 2007. http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5718a3.htm/
. Accessed February 19, 2016.
12. Al-Homood IA. Rheumatic conditions in patients with diabetes mellitus. Clin Rheumatol. 2013;32:527–533.
13. King KB, Rosenthal AK. The adverse effects of diabetes on osteoarthritis: update on clinical evidence and molecular mechanisms. Osteoarthritis Cartilage. 2015;23:841–850.
14. Liao KP, Solomon DH. Traditional cardiovascular risk factors, inflammation and cardiovascular risk in rheumatoid arthritis. Rheumatology (Oxford). 2013;52:45–52.
15. Morris CJ. Carrageenan-induced paw edema in the rat and mouse. Methods Mol Biol. 2003;225:115–121.
16. Stepanović-Petrović RM, Micov AM, Tomić MA, Ugrešić ND. The local peripheral antihyperalgesic effect of levetiracetam and its mechanism of action in an inflammatory pain model. Anesth Analg. 2012;115:1457–1466.
17. Vivancos GG, Verri WA Jr, Cunha TM, et al. An electronic pressure-meter nociception paw test for rats. Braz J Med Biol Res. 2004;37:391–399.
18. Fernández-Dueñas V, Sánchez S, Planas E, Poveda R. Adjuvant effect of caffeine on acetylsalicylic acid anti-nociception: prostaglandin E2 synthesis determination in carrageenan-induced peripheral inflammation in rat. Eur J Pain. 2008;12:157–163.
19. Hurley RW, Chatterjea D, Rose Feng M, Taylor CP, Hammond DL. Gabapentin and pregabalin can interact synergistically with naproxen to produce antihyperalgesia. Anesthesiology. 2002;97:1263–1273.
20. Rutten K, De Vry J, Robens A, Tzschentke TM, van der Kam EL. Dissociation of rewarding, anti-aversive and anti-nociceptive effects of different classes of anti-nociceptives in the rat. Eur J Pain. 2011;15:299–305.
21. Tomić MA, Micov AM, Stepanović-Petrović RM. Levetiracetam interacts synergistically with nonsteroidal analgesics and caffeine to produce antihyperalgesia in rats. J Pain. 2013;14:1371–1382.
22. Tallarida RJ. Drug Synergism and Dose-Effect Data Analysis. 2000.Boca Raton, FL: Chapman & Hall/CRC;
23. Tallarida RJ, Murray RB. Manual of Pharmacologic Calculations with Computer Programs. 1986.New York, NY: Springer Verlag;
24. Tallarida RJ, Stone DJ Jr, Raffa RB. Efficient designs for studying synergistic drug combinations. Life Sci. 1997;61:PL 417–425.
25. Tallarida RJ. The interaction index: a measure of drug synergism. Pain. 2002;98:163–168.
26. Yaksh TL, Noueihed RY, Durant PA. Studies of the pharmacology and pathology of intrathecally administered 4-anilinopiperidine analogues and morphine in the rat and cat. Anesthesiology. 1986;64:54–66.
27. Stepanović-Petrović RM, Micov AM, Tomić MA, Kovačević JM, Bošković BD. Antihyperalgesic/antinociceptive effects of ceftriaxone and its synergistic interactions with different analgesics in inflammatory pain in rodents. Anesthesiology. 2014;120:737–750.
28. Saisho Y. Metformin and inflammation: its potential beyond glucose-lowering effect. Endocr Metab Immune Disord Drug Targets. 2015;15:196–205.
29. Hyun B, Shin S, Lee A, et al. Metformin down-regulates TNF-α secretion via suppression of scavenger receptors in macrophages. Immune Netw. 2013;13:123–132.
30. Rocha AC, Fernandes ES, Quintão NL, Campos MM, Calixto JB. Relevance of tumour necrosis factor-alpha for the inflammatory and nociceptive responses evoked by carrageenan in the mouse paw. Br J Pharmacol. 2006;148:688–695.
31. Ji RR, Gereau RW IV, Malcangio M, Strichartz GR. MAP kinase and pain. Brain Res Rev. 2009;60:135–148.
32. Lisi L, Aceto P, Navarra P, Dello Russo C. mTOR kinase: a possible pharmacological target in the management of chronic pain. Biomed Res Int. 2015;2015:394257. doi:10.1155/2015/394257.
33. Tillu DV, Melemedjian OK, Asiedu MN, et al. Resveratrol engages AMPK to attenuate ERK and mTOR signaling in sensory neurons and inhibits incision-induced acute and chronic pain. Mol Pain. 2012;8:5. doi:10.1186/1744-8069-8-5.
34. Bas DB, Abdelmoaty S, Sandor K, et al. Spinal release of tumour necrosis factor activates c-Jun N-terminal kinase and mediates inflammation-induced hypersensitivity. Eur J Pain. 2015;19:260–270.
35. Galan A, Lopez-Garcia JA, Cervero F, Laird JM. Activation of spinal extracellular signaling-regulated kinase-1 and -2 by intraplantar carrageenan in rodents. Neurosci Lett. 2002;322:37–40.
36. Kwon SG, Roh DH, Yoon SY, et al. Blockade of peripheral P2Y1 receptors prevents the induction of thermal hyperalgesia via modulation of TRPV1 expression in carrageenan-induced inflammatory pain rats: involvement of p38 MAPK phosphorylation in DRGs. Neuropharmacology. 2014;79:368–379.
37. Norsted Gregory E, Codeluppi S, Gregory JA, Steinauer J, Svensson CI. Mammalian target of rapamycin in spinal cord neurons mediates hypersensitivity induced by peripheral inflammation. Neuroscience. 2010;169:1392–1402.
38. Obara I, Tochiki KK, Géranton SM, et al. Systemic inhibition of the mammalian target of rapamycin (mTOR) pathway reduces neuropathic pain in mice. Pain. 2011;152:2582–2595.
39. Sammons MJ, Raval P, Davey PT, Rogers D, Parsons AA, Bingham S. Carrageenan-induced thermal hyperalgesia in the mouse: role of nerve growth factor and the mitogen-activated protein kinase pathway. Brain Res. 2000;876:48–54.
40. 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.
41. Ortiz MI. Blockade of the antinociception induced by diclofenac, but not of indomethacin, by sulfonylureas and biguanides. Pharmacol Biochem Behav. 2011;99:1–6.
42. Dirig DM, Isakson PC, Yaksh TL. Effect of COX-1 and COX-2 inhibition on induction and maintenance of carrageenan-evoked thermal hyperalgesia in rats. J Pharmacol Exp Ther. 1998;285:1031–1038.
43. Guay J, Bateman K, Gordon R, Mancini J, Riendeau D. Carrageenan-induced paw edema in rat elicits a predominant prostaglandin E2 (PGE2) response in the central nervous system associated with the induction of microsomal PGE2 synthase-1. J Biol Chem. 2004;279:24866–24872.
44. Grond S, Sablotzki A. Clinical pharmacology of tramadol. Clin Pharmacokinet. 2004;43:879–923.
45. Hylden JL, Thomas DA, Iadarola MJ, Nahin RL, Dubner R. Spinal opioid analgesic effects are enhanced in a model of unilateral inflammation/hyperalgesia: possible involvement of noradrenergic mechanisms. Eur J Pharmacol. 1991;194:135–143.
46. Rezende RM, Dos Reis WG, Duarte ID, Lima PP, Bakhle YS, de Francischi JN. The analgesic actions of centrally administered celecoxib are mediated by endogenous opioids. Pain. 2009;142:94–100.
47. Whiteside GT, Boulet JM, Walker K. The role of central and peripheral mu opioid receptors in inflammatory pain and edema: a study using morphine and DiPOA ([8-(3,3-diphenyl-propyl)-4-oxo-1-phenyl-1,3,8-triaza-spiro[4.5]dec-3-yl]-acetic acid). J Pharmacol Exp Ther. 2005;314:1234–1240.
48. Jones CK, Eastwood BJ, Need AB, Shannon HE. Analgesic effects of serotonergic, noradrenergic or dual reuptake inhibitors in the carrageenan test in rats: evidence for synergism between serotonergic and noradrenergic reuptake inhibition. Neuropharmacology. 2006;51:1172–1180.
49. Stahl SM, Porreca F, Taylor CP, Cheung R, Thorpe AJ, Clair A. The diverse therapeutic actions of pregabalin: is a single mechanism responsible for several pharmacological activities? Trends Pharmacol Sci. 2013;34:332–339.
50. Fehrenbacher JC, Taylor CP, Vasko MR. Pregabalin and gabapentin reduce release of substance P and CGRP from rat spinal tissues only after inflammation or activation of protein kinase C. Pain. 2003;105:133–141.
51. Kumar N, Laferriere A, Yu JS, Leavitt A, Coderre TJ. Evidence that pregabalin reduces neuropathic pain by inhibiting the spinal release of glutamate. J Neurochem. 2010;113:552–561.
52. Okano K, Kuraishi Y, Satoh M. Involvement of spinal substance P and excitatory amino acids in inflammatory hyperalgesia in rats. Jpn J Pharmacol. 1998;76:15–22.
53. Yu LC, Hansson P, Brodda-Jansen G, Theodorsson E, Lundeberg T. Intrathecal CGRP8-37-induced bilateral increase in hindpaw withdrawal latency in rats with unilateral inflammation. Br J Pharmacol. 1996;117:43–50.
54. Graham GG, Punt J, Arora M, et al. Clinical pharmacokinetics of metformin. Clin Pharmacokinet. 2011;50:81–98.
55. Patrignani P, Patrono C. Cyclooxygenase inhibitors: from pharmacology to clinical read-outs. Biochim Biophys Acta. 2015;1851:422–432.
56. Reagan-Shaw S, Nihal M, Ahmad N. Dose translation from animal to human studies revisited. FASEB J. 2008;22:659–661.
57. Patrono C, Coller B, FitzGerald GA, Hirsh J, Roth G. Platelet-active drugs: the relationships among dose, effectiveness, and side effects: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 suppl):234S–264S.