Archives of Medical Research 44 (2013) 495e503

ORIGINAL ARTICLE Enhancement of Antinociception but not by Combinations Containing and in Arthritic Rats Francisco Javier Lopez-Mu noz,~ a Luis Alfonso Moreno-Rocha,a Guadalupe Bravo,a Uriah Guevara-Lopez, b Adriana Miriam Domınguez-Ramırez,c and Myrna Deciga-Camposd aDepartamento de Farmacobiologıa, Centro de Investigacion y Estudios Avanzados, Sede Sur, Mexico, D.F., Mexico bDireccion de Educacion e Investigacion en Salud, Unidad Medica de Alta Especialidad Dr. Victorio de la Fuente Narvaez, Instituto Mexicano del Seguro Social, Mexico, D.F., Mexico cDepartamento Sistemas Biologicos, Universidad Autonoma Metropolitana, Unidad Xochimilco, Mexico, D.F., Mexico dSeccion de Posgrado e Investigacion de la Escuela Superior de Medicina del Instituto Politecnico Nacional, Mexico, D.F., Mexico

Received for publication February 1, 2013; accepted August 26, 2013 (ARCMED-D-13-00072).

Background and Aims. The use of a combination of could provide an optimal treatment with minimal side effects. Combinations of tramadol and some nonste- roidal anti-inflammatory drugs have demonstrated synergistic antinociceptive effects as well as a significantly reduced occurrence of adverse effects. The purpose of this study was to investigate the antinociceptive and constipation effects of tramadol and metami- zole alone or in combination in rats and to discern among the types of drug interactions that exist using dose-response curves and an isobolographic analysis. Methods. The antinociceptive effects of tramadol and metamizole, alone or in various combination ratios, were quantitatively evaluated using the ‘‘pain-induced functional impairment model in the rat.’’ Additionally, the constipation effect was evaluated using the charcoal meal test. Results. Tramadol (3.2e56.2 mg/kg) and metamizole (56.2e562.3 mg/kg) demonstrated a dose-dependent response with tramadol being more efficacious and potent than meta- mizole. Twenty-five different combinations of tramadol with metamizole were analyzed, and the evaluated combinations exhibited antinociceptive effects that were either additive or potentiative. An optimal combination was established with 3.2 mg/kg of tramadol and 316.2 mg/kg of metamizole. However, the constipation observed with this combination was more severe than that observed with the administration of tramadol alone. Our results reveal a possible interaction between the two drugs, which may be pharmacokinetic and/ or pharmacodynamic in nature. Conclusions. The preclinical antinociceptive interaction and adverse effects produced by the combination of tramadol and metamizole suggests that caution should be exercised when us- ing this combination in the clinical therapy of pain. Ó 2013 IMSS. Published by Elsevier Inc. Key Words: Antinociception, Synergism, Constipation, Tramadol, Metamizole, PIFIR model.

Introduction using combination therapy is that, in some cases, the antino- ciceptive effects can be improved while the adverse effects Different combinations of and nonsteroidal anti- can be minimized. This postulate refers to the fact that inflammatory drugs (NSAIDs) are commonly used to con- improving antinociceptive effectiveness due to antinocicep- trol various pain statements (1). The potential advantage of tive synergism should enable a reduction in dosage and consequently a significant decrease in the incidence of ~ Address reprint requests to: Francisco Javier Lopez-Munoz, Departa- adverse effects (2). Previous studies with and mento de Farmacobiologıa, Centro de Investigacion y Estudios Avanzados, Sede Sur, Calzada de los Tenorios 235, Colonia Granjas Coapa, Mexico, metamizole (3) found an enhanced antinociceptive effect, D.F., C.P. 14330, Mexico; Phone: (þ52) (55) 5483-2851; FAX: (þ52) thus confirming the positive interaction between these two (55) 5483-2863; E-mail: fl[email protected] or fl[email protected] drugs. The efficacy of tramadol, which is a weak

0188-4409/$ - see front matter. Copyright Ó 2013 IMSS. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.arcmed.2013.09.004 496 Lopez-Mu noz~ et al./ Archives of Medical Research 44 (2013) 495e503 opioid-agonist with analgesic properties, is the result of a (Mexico City, Mexico), respectively. Uric acid was pur- complex interaction between opioid receptors with adren- chased from the Sigma Chemical Co. (St. Louis, MO) and ergic (4) and serotonergic systems (5). The combination of suspended in mineral oil. Either tramadol or metamizole tramadol with several NSAIDs, such as acetylsalicylic acid was administered subcutaneously (s.c.) using an isotonic (6), (7), (8), (9), saline solution as the vehicle (0.9% w/v). The drugs were (10) and (2) has shown a positive interaction in freshly prepared and administered at a volume of 2 mL/kg terms of its antinociceptive effects in a dose-dependent body weight for tramadol or metamizole. The doses manner. However, the combination of tramadol with rofe- mentioned in the text refer to the salts of these substances. coxib antagonized the antinociceptive effect of tramadol (11). Recently we demonstrated that, upon repeated admin- Measurement of Antinociceptive Activity istration, the positive synergistic antinociceptive effect of The antinociceptive activity was assessed using the PIFIR the combination of tramadol and metamizole generated anti- model, which has previously been described in detail nociceptive tolerance (12). Currently, there is little preclini- (13). The animals were lightly anesthetized with ethylic cal and quantitative evidence to establish the antinociceptive ether in an chamber, and was interaction of this particular combination of tramadol and induced by an intra-articular (i.a.) injection of 0.05 mL of metamizole. Consequently, the purpose of this study was 30% uric acid suspended in mineral oil in the knee joint to investigate the antinociceptive effects of tramadol and of the right hind limb. The intra-articular injection was per- metamizole, alone or in combination, using the ‘‘pain- formed through the patellar ligament using a 1 mL glass sy- induced functional impairment model in the rat (PIFIR ringe with a 24 gauge needle, 5 mm in length. Immediately model) (13) and the isobolographic analysis (14). The isobo- following this injection, an electrode was attached to the lographic method enables the discernment of the types of plantar surface of each hind paw (right and left) between interactions that are present and the classification of these ef- the plantar pads. The rats were then allowed to recover fects as potentiation, addition or sub-addition. Additionally, from anesthesia and placed on a 30-in diameter stainless the constipating effects produced by these drugs adminis- steel cylinder. The cylinder was rotated at 4 rpm, forcing tered alone or in combination were analyzed. the rats to walk. The variable to be measured was the time of contact between each hind paw and the cylinder. When the electrode placed on the animal’s paw made contact with Materials and Methods the cylinder floor, a circuit was closed, and the time the cir- Animals cuit remained closed was recorded. The cylinder was rotated for 2-min periods, during which the recordings were Male Wistar rats (Crl [WI]fBR) weighing 180e200 g from made; 28-min rest periods were interspersed between the the production unit of the laboratory of animal species of recordings. Following uric acid injection, the animals the Metropolitan Autonomous University were used. Ani- developed a progressive dysfunction of the injured limb. mals were housed in an acclimatized room (22 2 C) under The time of contact of the injured hind limb reached a zero a light-dark cycle of 12 h (lights on at 7:00), and experiments value 2.5 h after the uric acid injection. In all subsequent were performed during the light phase of the cycle. The experiments, treatments were administered 2.5 h after the experimental protocol (no. 0444-08) was approved by the uric acid injection. Thus, this time was considered to be local Institutional Animal Care and Use Committee in accor- time zero for the measurement of the antinociceptive effect, dance with the Mexican Federal Regulations for the Care and drugs were administered at this time. The time of con- and Use of Laboratory Animals NOM-062-ZOO-1999 tact was measured every 28 min for up to 4 h after drug (Mexican Ministry of Health). The protocol followed the administration. The data are expressed as a percentage of Guidelines on Ethical Standards for the investigation of the functionality index (FI%), i.e., the time of contact of Experimental Pain in Animals (15) and the Committee for the injected paw divided by the time of contact of the Research and Ethical Issues of the International Association control left paw x100. The recovery of the FI% was for the Study of Pain (16). At least 8 h before the experiment considered to be an expression of the antinociceptive effect. began, the rats were subjected to fasting with free access to Time-response curves were plotted to detect the onset of the drinking water. The number of animals used was kept to a antinociceptive effect, as were dose-response curves to minimum and the conditions were controlled. Each animal determine the antinociceptive efficacies and potency. All was used only once and, at the end of the study, the rats were experiments were performed between 7:00 and 14:00 h. euthanized in CO2 to avoid unnecessary . Measurement of Gastrointestinal Transit Drugs Gastrointestinal transit was assessed using the charcoal Tramadol hydrochloride and metamizole were obtained from meal test described previously by our group (17). Rats were Janssen-Cilag (Mexico City, Mexico) and Sanofi-Aventis s.c. administered the antinociceptive treatment (see study Antinociception/Constipation by Tramadol Plus Metamizole 497 design section) and, 30 min later, received 2 mL of a sus- AUC equivalent to the sum was expected. If the AUC of pension of 5% vegetable charcoal with 5% arabic gum in the tramadol þ metamizole combination was significantly a saline solution (0.9% w/v) via an intragastric flexible higher than the sum of the corresponding individual AUCs tube. Animals were euthanized in a CO2 chamber 30 min (theoretical sum), the result was considered to show poten- after the oral meal administration. Immediately, the stom- tiation. If it was similar to the theoretical sum, it was ach and small intestine were removed to measure the length considered to show an additive antinociceptive effect, but of the intestine (from the pyloric sphincter to the ileocecal if it was significantly lower than the theoretical sum, it junction) and the distance traveled by the charcoal meal. was considered to show a sub-additive interaction. To The propulsive activity of the gut was determined by calcu- determine the type of interaction, the values of the AUCs lating the percentage of gastrointestinal transit, i.e., the dis- were compared using an unpaired one-sided Student t-test tance travelled by the charcoal meal divided by the total (sum AUC expected vs. AUC obtained from the combina- length of the small intestine, x100. tion). The comparison of the FI% at two sampling times (Emax and E4h) was performed using a paired two-sided Study Design Student t-test. For the gastrointestinal transit data, one- way ANOVA for the comparison of two or more treat- The antinociceptive effects produced by tramadol and ments followed by Tukey’s multiple comparisons test metamizole administered alone or in combination were was used to detect significant differences; p #0.05 was evaluated in the PIFIR model. Independent groups of six considered to be the criterion for establishing significant rats were individually pretreated with an increasing 0.25 differences. logarithmic unit of tramadol (3.2, 5.6, 10, 17.8, 31.6 and Two isobolograms were constructed using ED and 56.2 mg/kg, s.c.) or metamizole (56.2, 100, 177.8, 316.2 30 ED (calculated from the maximal effect reached by each and 562.3 mg/kg s.c.) to obtain the corresponding dose- 50 compound) when the drugs were given alone or in combi- response curves. To analyze the possible synergistic inter- nation. To perform the isobolographic analysis, tramadol actions, 25 combinations of tramadol (3.2, 5.6, 10.0, 17.8 and metamizole were administered s.c. in combination as and 31.6 mg/kg) þ metamizole (56.2, 100, 177.8, 316.2 fixed ratio proportions of the equieffective ED (or and 562.3 mg/kg) were evaluated. Additionally, two groups 30 ED ) dose for tramadol and metamizole (1:1). The ED of six rats were s.c. administered the vehicle saline solution 50 30 (or ED ) value (SEM) for tramadol and metamizole (0.9% w/v). One group received the vehicle 2.5 h after the 50 alone were plotted on the x- and y-axes, respectively, uric acid (0.05 ml, i.a.) injection in the knee joint of the and the theoretical additive point was calculated according right hind limb and the other group received the vehicle to Tallarida (14). From the dose-response curve of the without the uric acid injection. combined drugs, the ED (or ED ) value of the total dose The intestinal transit of the charcoal meal was measured 30 50 of the combination was calculated. Statistical significance after the administration of the saline solution (vehicle of the difference between the theoretical additive point and group), tramadol (3.2 and 56.2 mg/kg s.c.) or the combina- the experimentally derived ED (or ED ) value was eval- tion of tramadol þ metamizole (3.2 þ 316.2 mg/kg s.c.). 30 50 uated using Student t-test. An experimental ED (or ED ) Based on previous studies, it is known that metamizole does 30 50 significantly smaller than the theoretical additive ED (or not affect the gastrointestinal transit (17); therefore, it was 30 ED ) was considered to indicate a potentiation ( p !0.05) decided that this drug would not be administered alone in 50 between tramadol and metamizole. the intestinal transit experiments.

Statistical Analysis Results Temporal curves were constructed by plotting the FI% against time for each treatment; antinociception was esti- The uric acid injection induced a complete dysfunction of mated as the recovery of the FI%, and the response for the right hind limb, corresponding to a FI% value of zero each treatment was presented as the mean SEM for in 2.5 h. Rats that did not receive uric acid, but received six animals. The cumulative antinociceptive effect during the vehicle alone (saline solution s.c.), did not show any the entire observation period (4 h) was determined as the change in the FI% during the observation period. Rats that area under the curve (AUC) of the temporal course to received uric acid and then the vehicle (saline solution s.c.) obtain the dose-response curve and to analyze the whole did not display any significant recovery of the FI% during antinociceptive effect elicited by the analgesic agent, either the observation period, i.e., the dysfunction of the right hind alone or in combination. The AUC was determined using limb was maintained throughout the experimental period, the method of trapezoids (18). The synergism between tra- which lasted another 4 h. At the doses used, tramadol and madol and metamizole was identified using dose-response metamizole did not affect the walking ability or produce curves and an isobolographic method (14). Based on the any motor impairment in the rats during the period of eval- additive effects of each individual drug component, an uation (data not shown). 498 Lopez-Mu noz~ et al./ Archives of Medical Research 44 (2013) 495e503

Antinociceptive Effects of the Drugs Assayed Individually The antinociceptive effect of tramadol or metamizole administered individually, which was measured as the AUC, increased in a dose-dependent manner, but displayed different efficacy. The dose-response curve for each anal- gesic drug is presented in Figure 1. Tramadol (56.2 mg/ kg) showed a significantly greater antinociceptive efficacy of 241.7 15.2 area units (au) than metamizole (562.3 mg/kg), which showed a maximum effect of 192.9 6.1 au ( p !0.05). From the calculation of the ED50 values (the estimated dose required to produce 50% of the maximum effect achievable in the model used or 187.5 au in the PIFIR model; the maximum value observed in this model was 375 au) for both drugs and it appears that trama- dol has a lower ED50 value (50.0 2.7 mg/kg) than meta- Figure 2. Antinociceptive effects obtained with metamizole (56.2, 100.0, 177.8, 316.2, or 562.3 mg/kg) alone or in combination with tramadol. mizole (559.2 22.9 mg/kg). The ED50 values of the drugs indicate that there was a significant difference in the antino- The y-axis represents the area under the curve of the time course ! (AUC); the x-axis depicts the doses (mg/kg) of metamizole administered ciceptive potency ( p 0.001) with tramadol being 11.2 simultaneously with tramadol, 3.2, 5.6, 10.0, 17.8 or 31.6 mg/kg. The high- times more potent than metamizole. There was no evidence est dashed line represents the maximum AUC value (375 area units) that of potential adverse effects at the doses used during the can be attained under these experimental conditions. The other two dashed experiment. lines represent the greatest change of antinociceptive effect, i.e., of meta- mizole 316.2 mg/kg alone (lower arrow) (the drug alone produced 109.7 9.4 au) to metamizole 316.2 mg/kg administered with tramadol 3.2 mg/kg Antinociceptive Effects of the Drug Combinations (the AUC produced was 307.3 10.9 au). The upper arrow shows how the The antinociceptive effects obtained with the analgesic maximum antinociceptive efficacy of metamizole 562.3 mg/kg changes from 192.9 6.1 to 329.9 7.8 au when metamizole (562.3 mg/kg) drugs administered individually or in combination in was administered with tramadol 3.2 mg/kg. The study showed that 22 com- different proportions are shown in Figure 2. This graph binations presented various degrees of potentiation, whereas the others ex- was constructed using the mean from six animals for each hibited only additive antinociceptive effects. Each point represents the dose either alone or in combination. The x-axis shows the mean of six experiments SEM. doses either of metamizole or tramadol administered alone or simultaneously. The y-axis corresponds to the AUC effect during a 4 h observation period) of the drugs admin- value of the temporal courses (the overall antinociceptive istered alone or in combination. The dose-response curve () corresponds to the antinociceptive effects produced by metamizole administered alone; at 562.3 mg/kg this curve showed a maximum effect of 192.9 6.1 au. When trama- dol was administered alone at 31.6 mg/kg, an antinocicep- tive effect of 63.2 19.5 au was shown. As shown in Figure 2, the 25 combinations of tramadol with metamizole used in this study produced much better antinociceptive ef- fects than the individually administered compounds. The results from the statistical comparison indicate a significant interaction. The maximal antinociceptive effect produced by the tramadol and metamizole combination (3.2 and 562.3 mg/kg, respectively) was 329.9 7.8 au. In all of the combinations tested the interaction between both com- pounds is positive. The difference between the effect of the combination and the sum of the individual effects were calculated. The results significantly higher than level ‘‘0’’ were considered Figure 1. Comparison of the dose-response curves of tramadol (s.c.) and to indicate potentiation, whereas the other results were metamizol (s.c.) in the pain-induced functional impairment model in the considered to indicate addition. Locating the points that rat. The antinociceptive response is expressed on the y-axis as the area un- showed significant statistical improvement could be useful der the curve (AUC) of the functionality index over the 4 h observation period. It can be seen that, at the maximum dose, tramadol showed a higher for distinguishing between potentiation and additive effects. þ maximum effect and a lower ED50 than metamizole. Data are expressed as Thus, three combinations of tramadol metamizole pro- the mean SEM for six animals. duced additive antinociceptive effects, and 22 combinations Antinociception/Constipation by Tramadol Plus Metamizole 499 produced potentiation effects. There was no antagonistic ef- (A: 3.2 mg/kg tramadol þ 316.2 mg/kg metamizole) and fect observed in the combinations tested in this study. The by the combination that produced the maximum antinoci- combinations that produced only additive antinociceptive ceptive effect (B: 3.2 mg/kg tramadol and 562.3 mg/kg effects were 5.6 þ 100, 10 þ 56.2 and 17.8 þ 56.2 mg/ metamizole) are shown in Figure 3. kg of tramadol þ metamizole. Various degrees of potentia- Another approach for investigating the interaction tion effects were observed and the combinations that between the two selected analgesic drugs is the isobolo- displayed the most potentiation are listed in Table 1. graphic method (14). Isobolographic analysis using fixed Using this graph, it is easy to visualize the type of drug ratio (1:1) ED30 and ED50 revealed a significant potentia- interaction of tramadol and metamizole (i.e., addition or tive interaction between tramadol and metamizole after potentiation). However, the combinations that produce s.c. administration in the PIFIR model (Figure 4). Horizon- potentiation and using the lower concentration of tramadol tal and vertical bars indicate SEM. The oblique line could be more promising because fewer adverse effects can between the x- and y-axes is the theoretical additive line. be expected. Two doses of tramadol, 3.2 and 5.6 mg/kg The point for experimental ED30 (or ED50) for this combi- when combined with 316.2 mg/kg of metamizole, showed nation was obtained below the theoretical additive line indi- significantly increased potentiation (307.3 10.9 and cating potentiative interaction ( p !0.05). Further, the 311.0 13.6 au, respectively). At a dose of 3.2 mg/kg, tra- experimental ED30 (or ED50) dose was significantly smaller madol alone did not produce any antinociceptive effect (0.0 than the calculated additive ED30 (or ED50) doses, thereby 0.0 au), and at the dose of 316.2 mg/kg, metamizole demonstrating synergistic interaction. alone led to an AUC of 109.7 9.4 au; however, the com- bination of tramadol þ metamizole (3.2 þ 316.2 mg/kg) Effect of the Tramadol þ Metamizole Combination on the produced an AUC of 307.3 10.9 au, which is significantly Gastrointestinal Transit higher than the expected AUC resulting from the sum of the individual values (i.e., 109.7 9.4 au) ( p !0.001). The The percentages of the gastrointestinal transit of the char- Emax data from the corresponding temporal course were coal meal after a single dose of vehicle (saline solution 0.0 0.0 and 64.9 13.1 for tramadol and metamizole, 0.9%) and after the combination that produced a high respectively. Analysis of the Emax data showed an increase potentiation of the antinociceptive effect (3.2 þ 316.2 in the values obtained from the combination (91.3 2.7%), mg/kg, tramadol þ metamizole), as well as after tramadol which were significantly higher than the sum of the individ- administered individually at doses of 3.2 mg/kg and 56.2 ual effects (64.9 13.1%). Other examples of potentiation mg/kg (which present the maximum antinociceptive effect), with tramadol þ metamizole are shown in Table 1. The are shown in Figure 5. For treatment with the vehicle alone, temporal courses of the antinociceptive effects produced the charcoal meal traveled 49.5 4.5% of the total length by the combination that resulted in high potentiation of the small intestine in 30 min. The gastrointestinal transit

Table 1. Antinociceptive effects produced by some combinations and those produced by the same doses of the compounds administered individually, tramadol (TRA) or metamizole (MET)

a b c Treatment Dose (mg/kg) AUC (au) Emax (FI%) TEmax (h) E4h (FI%)

TRA 3.2 0.0 0.0 0.0 0.0 - 0.0 0.0 MET 562.3 192.9 4.3 88.6 8.1 1.0 10.1 5.8 TRA þ MET (1) 329.9 7.8* 90.2 3.9 1.5 88.2 3.4 TRA 10.0 1.3 0.4 1.6 0.7 0.5 0.0 0.0 MET 562.3 192.9 4.3 88.6 8.1 1.0 10.1 5.8 TRA þ MET (2) 313.5 24.1* 88.2 4.8 2.5 87.3 4.0 TRA 31.6 63.2 19.5 33.3 13.5 2.0 23.0 13.3 MET 562.3 192.9 4.3 88.6 8.1 1.0 10.1 5.8 TRA þ MET (3) 313.5 24.1* 88.2 4.8 2.5 87.3 4.0 TRA 3.2 0.0 0.0 0.0 0.0 - 0.0 0.0 MET 316.2 109.7 9.4 64.9 13.1 0.5 2.0 1.2 TRA þ MET (4) 307.3 10.9* 91.3 2.7 1.0 76.6 6.7 TRA 5.6 0.0 0.0 0.0 0.0 - 0.0 0.0 MET 316.2 109.7 9.4 64.9 13.1 0.5 2.0 1.2 TRA þ MET (5) 311.0 13.6* 89.8 5.4 1.5 76.5 7.2

Note: 1e3 are combinations that produced maximum antinociceptive effect; 4 and 5 are combinations that produced higher potentiation. *p !0.01. aArea under the curve of the time course. bTime required to produce the maximum effect in the time course curve. cAntinociceptive effect obtained 4 h after the administration. 500 Lopez-Mu noz~ et al./ Archives of Medical Research 44 (2013) 495e503

Figure 3. Temporal courses of the combinations that produced (A) high potentiation: 3.2 mg/kg tramadol (B), 316.2 mg/kg metamizole (-) and the combination of tramadol þ metamizole (3.2 þ 316.2 mg/kg) (D) and (B) the maximum antinociceptive effect: 3.2 mg/kg tramadol (B), 562.3 mg/kg metamizol (-) and the combination of tramadol þ metamizole (3.2 þ 562.3 mg/kg) (D). The first combination represents Figure 4. Isobolograms describing the synergistic interaction between one of the two combinations that showed the highest potentiation of the tramadol and metamizol using ED30 (4A) and ED50 (4B). Horizontal antinociceptive effects. The AUC (307.3 10.9 au) obtained with this and vertical bars indicate SEM. The oblique line that connects the individ- combination was significantly higher ( p !0.01) than the global effect of ual ED30 values in each combination (-) is the theoretical additive line. the sum of the individual effects (109.7 9.4 au). Data are expressed as The point in this line is the theoretical additive point (B) calculated from the mean (n 5 6) SEM. the individual drug ED30 (or ED50) values. In all cases, the experimental ED30 and ED50 values (C) lie far below the additive lines indicating a significant synergism. *p !0.05. at a dose of 3.2 mg/kg of tramadol (42.17 3.8%) was not significantly different from that obtained with the saline so- has high sensitivity (13). Several clinical (19) and experi- lution. However, treatment with tramadol at a dose of 56.2 mental (2,6) evidences support the potential antinociceptive mg/kg and the combination of tramadol þ metamizole effect of combining tramadol with some NSAIDs. Several of (3.2 þ 316.2 mg/kg) showed intestinal transits of 20.8 these combinations have proven to be more effective than 5.8% and 2.8 1.4%, respectively; both responses were either drug administered separately. In the present study, significantly different from that of the control group. dose-response curves analysis was conducted, which has previously been used to determine positive or negative inter- actions between several different drug combinations such as Discussion morphine þ metamizole (3), morphine þ acetylsalicylic This is the first study analyzing the antinociceptive effects of acid (20), D-propoxyphene þ acetylsalicylic acid (21), combinations of tramadol and metamizole using the PIFIR D-propoxyphene þ acetaminophen (22), and tramadol þ model and the isobolographic analysis. The PIFIR model ketorolac (2); all these combinations were found to produce was used because it enables the evaluation of the temporal different degrees of antinociceptive potentiation. The dose- course of the antinociceptive effect in the same animal; response curve analysis has demonstrated an antagonism furthermore, it does not generate conditioned learning and between the combination of tramadol and (11). Antinociception/Constipation by Tramadol Plus Metamizole 501

pronounced than that observed with a higher dose in the dose-response curve of tramadol (56.2 mg/kg). These results were somewhat unexpected because it was previously re- ported that, for the association between an opioid compound (morphine) and a NSAID (metamizole), the constipation ef- fect was not as reinforced as the antinociceptive effect (17). From the results obtained in the present study, we can infer that the combination of an opioid with a NSAID such as metamizole does not always result in the absence or reduc- tion of its adverse effects, and that this effect may vary depending on the nature of the opioid compound. Previous studies showed that a low dose of metamizole (85 mg/kg) combined with a gradual increase in the dosage of tramadol increased the intestinal motility compared with the administration of tramadol alone (25). This particular Figure 5. Effects of the single administration of tramadol (3.2 and 56.2 report, which is apparently contradictory to our present re- mg/kg, s.c.) and its combination with 316.2 mg/kg metamizole on the sults, showed that metamizole and other NSAIDs antago- gastrointestinal transit of charcoal meal. Bars are the mean SEM of nize the constipating effects, depending on the individual six individual readings. *Significantly different from the saline treatment as determined by one-way ANOVA followed by Tukey’s multiple compar- compounds. In our study, the dose of metamizole was high- isons test ( p !0.05). er (316.2 mg/kg) than that tested by Planas and coworkers, indicating that the effect of constipation depends on the magnitude of the administered dose of metamizole. The application of a dose-response curves methodology per- In addition to changes in gastrointestinal motility, the mits the characterization of the full dose-response relation results showing the antinociceptive synergy between trama- and can therefore be used to develop practical guidelines dol and metamizole clearly indicate that there is an interac- for optimal drug dosing. Additionally, the use of dose- tion between the drugs. This interaction may involve response curves instead of a single point approach has been changes at both the pharmacodynamic (PD) and/or pharma- recommended (23). Other types of interaction analyses such cokinetic (PK) levels. However, in the current study we did as the method of the isobolograms are also used and our not endeavor to determine whether PD or PK changes group has published some drug interaction studies in which occurred nor did we aim to understand the mechanisms we conducted both analyses (2,11). In the present study, of action involved when these drugs were administered because the degree of potentiation of the antinociceptive together. The combination of centrally and peripherally effect depends on the dose rate of the combination, the deci- acting analgesics has theoretical justification because these sion to employ both analyses was made. Our aim was to compounds work through different mechanisms. Tramadol, identify the combination that would result in the highest which is an atypical and synthetic opioid analgesic, is a degree of potentiation and the combination that would pro- centrally acting agent whose mechanisms of action is not duce the greatest degree of relief. completely understood. There is evidence that this com- Of 25 combinations tested, 22 showed effects of potenti- pound exhibits only a low affinity for opioid receptors ation (different degrees of potentiating effects), whereas the and lacks selectivity for m-, k-, or d-binding sites (26,27). remaining combinations showed additive effects. This poten- The analgesic efficacy of tramadol is the result of a com- tiating effect of tramadol and metamizole was observed in plex interaction between opioid receptors with adrenergic acetic acid-induced nociception (24) and in the plantar test (4) and serotonergic systems (5). Metamizole is a centrally (12). Using the dose-response curves, the combination with and peripherally acting agent that exerts its analgesic ef- the highest degree of potentiation but with the lowest dose fects through several mechanisms, of which the best of opioid drug was selected (3.2 þ 316.2 mg/kg tramadol described ones are cyclooxygenase inhibition (28), delayed þ with metamizole) in order to establish the possible changes activation of the L-arginine/nitric oxide/cGMP/K channel in constipation as an adverse effect, and it is present along pathway (29), activation of the descending inhibitory pain with the antinociceptive effect. This particular drug dose control system (30), interaction with the glutamatergic sys- combination could be expected to have a low incidence of tem (31), and release of endogenous opioid peptides (32). adverse effects. However, this combination significantly Additionally, with regard to gastrointestinal effects and on reduced intestinal motility, which was comparable to that the basis of the results observed in this study, higher doses seen with the administration of morphine alone, as reported of metamizole could result in a reduced intestinal transit, in previous studies (17). This decrease was greater than that more than that with the individual administration of trama- presented by the administration of tramadol alone at the same dol. This result could partly be attributed to the dose as that in combination (3.2 mg/kg) and was even more metamizole-induced production of endogenous , 502 Lopez-Mu noz~ et al./ Archives of Medical Research 44 (2013) 495e503 which act similarly to exogenously administered opioids, that changes in the drug dosage in combination therapy especially with respect to the adverse effects and binding can change the degree of interaction and the degree of anal- to m receptors in the gastrointestinal tract (33). These results gesic synergism that occurs, c) when two analgesics are are also consistent with previous studies that showed that combined, analgesic synergism is produced, and we must the direct administration of metamizole to the gray peri- ensure that adverse effects are not enhanced, and d) anal- aqueductal mater or the systemic administration of metami- ysis of complete CDR combinations is possible using labo- zole (34) activates endogenous opioid circuits along the ratory animals, but not clinically. Hence, these studies descending control path (30,32). provide relevant information that may be useful now and Thus, the PD mechanism for the interaction between can be confirmed later in clinical studies. tramadol and metamizole in the gastrointestinal effect In summary, the current data support previous results that could be attributed partially to their participation in showed that acute treatment with a combination of opioid and the opioidergic system. Other mechanisms such as the L- NSAID drugs is effective in improving antinociceptive ef- arginine-NO-cyclic GMP pathway and interaction with fects, although this is the first study using the PIFIR model N-methyl D-aspartic acid receptors could be proposed to and dose-response curves. In this study, several tramadol þ explain the antinociceptive synergism observed with the metamizole combinations demonstrated potentiation of anti- combination of such drugs. However, a PK interaction be- nociceptive effects but also promoted an increase in side ef- tween tramadol and metamizole cannot be disregarded fects associated with intestinal motility. The clinical and warrants further investigation. implications of this study are important given the desire to The marketed tramadol is the racemate of the trans- maximize analgesia while minimizing the adverse effects isomers. PK studies have shown that tramadol is rapidly in a variety of situations in which pain is implicated. and almost completely absorbed after administration, but its mean absolute bioavailability is only 65e70% due to the first-pass hepatic metabolism (35). Tramadol has high tis- Acknowledgments sue affinity, and its plasma protein binding is |20% (36). This We wish to thank L. Oliva and F. Sanchez for technical assistance. analgesic is rapidly and extensively metabolized in the liver, L.A. Moreno-Rocha is a fellow of the National Council for Sci- and the principal metabolic pathways, O- and N-demethyla- ences and Technology (CONACYT), Mexico. tion, involve cytochrome P-450 (37). The primary metabo- lites, O-desmethyltramadol (pharmacologically active) and N-desmethyltramadol, may be further metabolized to three References additional secondary metabolites. The O-demethylated me- 1. Schug SA. Combination analgesia in 2005—a rational approach: focus tabolites are conjugated with glucuronic acid before excre- on paracetamol-tramadol. Clin Rheumatol 2006;25:S16eS21. ~ tion into the urine (35). On the other hand, metamizole is a 2. Lopez-Munoz FJ, Dıaz-Reval MI, Terron JA, et al. Analysis of the analgesic interactions between ketorolac and tramadol during arthritic pro-drug that undergoes hydrolysis to form 4-methyl- nociception in rat. Eur J Pharmacol 2004;484:157e165. amino-antipyrine (pharmacologically active), which is then 3. Lopez-Mu noz~ FJ. Surface of synergistic interaction between dipyrone converted to 4-formyl-amino-antipyrine and 4-amino-anti- and morphine in the PIFIR model. Drug Dev Res 1994;33:26e32. pyrine (pharmacologically active). The latter is then acety- 4. Kayser V, Besson JM, Guilbaud G. Evidence for a noradrenergic lated to form 4-acetylaminoantipyrine. Approximately 20% component in the antinociceptive effect of the analgesic agent trama- dol in an animal model of clinical pain, the arthritic rat. Eur J Pharma- of 4-methyl-amino-antipyrine is eliminated as the glucuro- col 1992;224:83e88. nide (38,39). Therefore, it is possible that tramadol and meta- 5. Oliva P, Aurilio C, Massimo F, et al. The antinociceptive effect of tra- mizole could compete for the same enzymes, causing madol in the formalin test is mediated by the serotonergic component. changes in the concentrations of tramadol and its metabolites Eur J Pharmacol 2002;445:179e185. ~ and, consequently, in the pharmacological effects produced. 6. Salazar LA, Ventura-Martınez R, Lopez-Munoz FJ. Synergistic anti- nociceptive interaction between and tramadol, the atypical It will be very interesting to confirm with reports of opioid analgesic in the rat. Drug Dev Res 1995;36:119e224. molecular biomarkers the relation between metamizole 7. Zhang Y, Du L, Pan H, et al. Enhanced analgesic effects of propace- and endogenous opioids. ‘‘The central effects of metami- tamol and tramadol combination in rats and mice. Biol Pharm Bull zole have been associated with the endogenous opioid sys- 2011;34:349e353. tem because its antinociceptive effect is blocked by 8. Striebel HW, Hackenberger J. A comparison of a tramadol/metamizole infusion with the combination tramadol infusion plus ibuprofen sup- naloxone (40,41) and there is evidence that this drug is able positories for postoperative following hysterectomy. to stimulate the release of pituitary and hypothalamic b- Anaesthesist 1992;41:354e360. endorphins’’ (42). 9. Satyanarayana PS, Jain NK, Singh A, et al. Isobolographic analysis of Finally, our study highlighted the following: a) it is interaction between cyclooxygenase inhibitors and tramadol in acetic important to understand the type of interaction that can acid-induced writhing in mice. Prog Neuropsychopharmacol Biol Psy- chiatry 2004;28:641e649. be found with different analgesics for one particular type 10. Lauretti GR, Mattos AL, Lima IC. Tramadol and beta-cyclodextrin of pain because this type of interaction can be altered by piroxicam: effective multimodal balanced analgesia for the intra- manipulating the type of pain, b) it is important to know and postoperative period. Reg Anesth 1997;22:243e248. Antinociception/Constipation by Tramadol Plus Metamizole 503

11. Garcıa-Hernandez L, Deciga-Campos M, Guevara-Lopez U, et al. tramadol, an ‘‘atypical’’ opioid analgesic. J Pharmacol Exp Ther Co-administration of rofecoxib and tramadol results in additive or 1992;260:275e285. sub-additive interaction during arthritic nociception in rat. Pharmacol 28. Campo C, de Gregorio R, Garcıa-Nieto R, et al. Regulation of cyclo- Biochem Behav 2007;87:331e340. oxygenase activity by metamizol. Eur J Pharmacol 1999;335: 12. Moreno-Rocha LA, Domınguez-Ramırez AM, Cortes-Arroyo AR, 154e159. et al. Antinociceptive effects of tramadol in co-administration with 29. Alves DP, Duarte IDG. Involvement of ATP-sensitive Kþ channels in metamizol after single and repeated administrations in rats. Pharmacol the peripheral antinociceptive effect induced by dipyrone. Eur J Phar- Biochem Behav 2012;103:1e5. macol 2002;444:47e52. 13. Lopez-Mu noz~ FJ, Salazar LA, Castaneda-Hern~ andez G, et al. A new 30. Hernandez N, Vanegas H. Antinociception induced by PAG- model to assess analgesic activity: pain-induced functional impair- microinjected dipyrone (metamizol) in rats: involvement of spinal ment in the rat (PIFIR). Drug Dev Res 1993;28:169e175. endogenous opioids. Brain Res 2001;896:175e178. 14. Tallarida RJ. Statistical analysis of drug combinations for synergism. 31. Beirith A, Santos ARS, Rodrıguez ALS, et al. Spinal and supraspinal Pain 1992;49:93e97. antinociceptive action of dipyrone in formalin, and gluta- 15. Zimmermann M. Ethical guidelines for investigations of experimental mate tests. Study of the mechanism of action. Eur J Pharmacol pain in conscious animals. Pain 1983;16:109e110. 1998;345:233e245. 16. Covino BG, Dubner R, Gybels J, et al. Ethical standards for investiga- 32. Vasquez E, Vanegas H. The antinociceptive effect of PAG- tions of experimental pain in animals. The Committee for Research microinjected dipyrone in rats is mediated by endogenous opioids of and Ethical issues of the International Association for the Study of the rostral ventromedical medulla. Brain Res 2000;854:249e252. Pain. Pain 1980;9:141e143. 33. Akman H, Aksu F, Gultekin€ I, et al. A possible central antinociceptive 17. Hernandez-Delgadillo GP, Ventura MR, Dıaz RMI, et al. Metamizol effect of dipyrone. Pharmacology 1996;53:71e78. potentiates morphine antinociception but not constipation after 34. Vazquez E, Hernandez N, Escobar W, et al. Antinociception induced chronic treatment. Eur J Pharmacol 2002;441:177e183. by intravenous dipyrone (metamizol) upon dorsal horn neurons: 18. Gibaldi M. Estimation of area under the curve. In: Gibaldi M, ed. Bio- involvement of endogenous opioids at the periaqueductal gray matter, pharmaceutics and Clinical Pharmacokinetics. 4th ed. Philadelphia: the nucleus raphe magnus and the spinal cord in rats. Brain Res 2005; Lea & Febiger; 1991. pp. 377e378. 1048:211e217. 19. Pergolizzi JV, van de Laar M, Langford R, et al. Tramadol/paraceta- 35. Ardakani YH, Mohammad-Reza R. Pharmacokinetics of tramadol and mol fixed-dose combination in the treatment of moderate to severe its three main metabolites in healthy male and female volunteers. Bio- pain. J Pain Res 2012;5:327e346. pharm Drug Dispos 2007;28:526e533. 20. Lopez-Mu noz~ FJ, Villalon CM, Terron JA, et al. Doses of acetyl sal- 36. Lee CR, McTavish D, Sorkin EM. Tramadol: a preliminary review icylic acid and morphine in combination which provided either of its pharmacodynamic and pharmacokinetic properties, and thera- maximal levels of analgesia or the highest potentiation effect in the peutic potential in acute and states. Drugs 1993;46: rat. Drug Dev Res 1995;35:94e101. 313e340. 21. Lopez-Mu noz~ FJ. Profile of analgesic interaction between aspirin and 37. Lintz W, Barth H, Becker R, et al. Pharmacokinetics of tramadol and d-propoxyphene obtained by means of the surface of synergistic inter- bioavailability of enteral tramadol formulations. 2nd communication: action. Drug Dev Res 1995;35:13e19. drops with ethanol. Arznemittelforschung 1998;48:436e447. 22. Lopez-Mu noz~ FJ, Salazar LA. Determination of analgesic interaction 38. Domınguez-Ramırez AM, Cortes-Arroyo AR, Hurtado M, et al. Effect between acetaminophen and d-propoxyphene obtained by means of of metamizol on morphine pharmacokinetics and pharmacodynamics the surface of synergistic interaction. Methods Find Exp Clin Pharma- after acute and subchronic administration in arthritic rats. Eur J Phar- col 1995;17:311e320. macol 2010;645:94e101. 23. Sawynok J, Yaksh TL. as an analgesic adjuvant: a review of 39. Bacracheva N, Kamali F. Lack of interaction between zidovudine and pharmacology and mechanisms of action. Pharmacol Rev 1993;45: 4-methyl-amino-antipyrine, the active metabolite of metamizole, in 43e85. human liver in vitro. Methods Find Exp Clin Pharmacol 1995;17: 24. Poveda R, Planas E, Pol O, et al. Interaction between metamizol and 299e301. tramadol in a model of acute visceral pain in rats. Eur J Pain 2003;7: 40. Dogrul A, Gulmez SE, Deveci MS, et al. The local antinociceptive ac- 439e448. tions of nonsteroidal antiinflammatory drugs in the mouse radiant heat 25. Planas E, Poveda R, Sanchez S, et al. Non-steroidal anti-inflammatory tail-flick test. Anesth Analg 2007;104:927e935. drugs antagonise the constipating effects of tramadol. Eur J Pharmacol 41. Tortorici V, Vazquez E, Vanegas H. Naloxone partial reversal of the 2003;482:223e226. antinociception produced by dipyrone microinjected into the periaque- 26. Hennies HH, Friderichs E, Schneider J. Receptor binding, analgesic ductal gray of rats. Possible involvement of medullary off- and on- and antitussive potency of tramadol and other selected opioids. Arz- cells. Brain Res 1996;725:106e110. neimittelforschung 1988;38:877e881. 42. Nikolova I, Tencheva J, Voinikov J, et al. Metamizole: a review profile 27. Raffa RB, Friderichs E, Reimann W, et al. Opioid and monopioid of a well-known ‘‘forgotten’’ drug. I: Pharmaceutical and nonclinical components independently contribute to the mechanism of action of profile. Biotechnol Biotechnol Equiv 2012;26:3329e3337.