Clinical effects and pharmacokinetic variables of romifidine and the peripheral a 2 -adrenoceptor antagonist MK-467 in horses

Veterinary Anaesthesia and Analgesia, 2016, 43, 599 610 doi:10.1111/vaa.12354 RESEARCH PAPER Clinical effects and pharmacokinetic variables of romifidine and the peripheral a 2 -adrenoceptor antagonist MK-467 in horses Annemarie de Vries*,, Soile AE Pakkanen, Marja R Raekallio, Abel Ekiri, Mika Scheinin,, Polly M Taylor** & Outi M Vainio *Davies Veterinary Specialists, Higham Gobion, UK Department of Equine and Small Animal Medicine, University of Helsinki, Helsinki, Finland College of Public Health and Health Professions, University of Florida, Gainesville, FL, USA Department of Pharmacology, Drug Development and Therapeutics, University of Turku, Turku, Finland Unit of Clinical Pharmacology, Turku University Hospital, Turku, Finland **Taylor Monroe, Little Downham, UK Correspondence: Annemarie de Vries, Davies Veterinary Specialists, Manor Farm Business Park, Higham Gobion, Hitchin SG5 3HR, UK. E-mail: Marieke.devries@vetspecialists.co.uk Abstract Objectives To investigate the effects of MK-467 on sedation quality, and cardiopulmonary and pharmacokinetic variables in horses sedated intravenously (IV) with romifidine. Study design Experimental, randomized, crossover design. Animals Seven healthy mares. Methods Romifidine (80 lg kg 1 ; R) and MK-467 (200 lg kg 1 ; MK) were administered IV alone and in combination (R + MK). Levels of sedation and borborygmi were scored. Heart rate (HR), direct arterial blood pressure (ABP) and respiratory rate (f R ) were recorded. Arterial and venous blood gas analyses were performed and venous plasma drug concentrations were measured. Pharmacokinetic parameters were calculated. Linear mixed modelling for repeated measures, contrasts of least square means by Bonferroni correction tests, one-way ANOVA for repeated measures with Bonferroni multiple comparison tests and paired Student s t-tests were used to compare results within and between treatments as appropriate. Significance was set at p < 0.05. Results After R, ABP increased and HR and f R decreased significantly. After R + MK, HR, f R, systolic and mean ABP decreased. MK alone increased both HR and f R. After R, ABP was significantly higher than after R + MK. HR and f R were significantly higher after MK than after R and R + MK. Areas under the curve for sedation time were similar after R and R + MK. Intestinal activity decreased markedly after R and less after R + MK. Volume of distribution and clearance of romifidine were significantly higher and area under the concentration time curve extrapolated to infinity significantly lower after R + MK than after R. Conclusions Combined romifidine and MK-467 prevented the cardiovascular changes commonly seen with romifidine but did not affect sedation quality. Clinical relevance Combined IV romifidine and MK- 467 can be used to attenuate the cardiovascular effects of romifidine, such as in horses with colic or undergoing general anaesthesia. 599

Keywords horses, MK-467, peripheral antagonist, sedation, a 2 -agonist. Introduction Romifidine, an a 2 -adrenoceptor agonist and imidazole derivative, is widely used for sedation and preanaesthetic medication in horses. In addition to desirable sedative and analgesic effects, its administration results in characteristic cardiovascular changes, including increased systemic vascular resistance (SVR), reflex bradycardia, cardiac conduction disturbances, reduced cardiac output (CO) and increased central venous pressure (CVP) (Clarke et al. 1991; England et al. 1992; Freeman & England 2000; Freeman et al. 2002; Figueiredo et al. 2005; Wojtasiak-Wypart et al. 2012). Along with these well-known side effects, a 2 -adrenoceptor agonists also reduce gut motility (Roger & Ruckebusch 1987; Freeman & England 2001; Elfenbein et al. 2009; Mama et al. 2009; Zullian et al. 2011; Vainionp a a et al. 2013), which should ideally be avoided in horses, a species prone to ileus and impaction. The combined administration of a peripherally restricted a 2 -antagonist with an a 2 -agonist such as romifidine could be a useful pharmacological strategy to facilitate the advantageous sedative, analgesic and central sympatholytic properties of romifidine while avoiding its potentially detrimental haemodynamic and gastrointestinal side effects. MK-467 (previously also known as L-659,066) is a peripherally acting selective a 2 -antagonist. Because of its low lipid solubility, its blood brain barrier penetration after systemic administration is poor (Clineschmidt et al. 1988). The concomitant administration of MK-467 attenuated the cardiovascular side effects of the a 2 -agonists medetomidine and dexmedetomidine in dogs (Enouri et al. 2008; Honkavaara et al. 2011; Rolfe et al. 2012; Salla et al. 2014) and sheep (Bryant et al. 1998; Raekallio et al. 2010) and had minimal effects on medetomidine- (Rolfe et al. 2012) and dexmedetomidine-induced (Honkavaara et al. 2008; Restitutti et al. 2011) sedation in dogs. In horses, its simultaneous administration with detomidine, another a 2 -agonist, has been demonstrated to either abolish or at least attenuate some of the peripherally induced cardiovascular effects caused by detomidine without affecting clinical sedation (Vainionp a a et al. 2013). Compared with detomidine alone, the concomitant use of MK-467 with detomidine for pre-anaesthetic medication resulted in enhanced cardiac function and tissue oxygen delivery in horses under isoflurane-maintained anaesthesia (Pakkanen et al. 2015). In ponies, administration of MK-467 before medetomidine reduced the medetomidineinduced cardiovascular effects (Bryant et al. 1998). No studies have been published on the effects of the combined use of MK-467 with romifidine in horses. Furthermore, only one published study, in four ponies, mentions the effects of MK-467 alone [264 lg kg 1 intravenously (IV)] on cardiovascular indices (Bryant et al. 1998). No effect on heart rate (HR) and no significant effect on arterial blood pressure (ABP) were noted, although exact values for these variables were not reported. Infusion of MK-467 to human volunteers did not affect either HR or ABP (Sciberras et al. 1994). An increase in HR, cardiac index, oxygen consumption, oxygen delivery and a decrease in SVR occurred in dogs after IV injection of 200 lg kg 1 of MK-467 (Enouri et al. 2008). Honkavaara et al. (2011) reported an increase in cardiac index and tissue oxygen delivery after an IV dose of 250 lg kg 1 in the same species. The characteristic haemodynamic effects induced by a 2 -agonists may impair the systemic clearance of concomitantly administered drugs because of decreased hepatic blood flow and consequently reduced drug metabolism; changes in tissue perfusion may also result in alterations in drug distribution. Atipamezole, an a 2 -antagonist, increased the elimination of medetomidine in dogs, probably by restoring hepatic blood flow (Salonen et al. 1995). Similarly, MK-467 has been shown to reduce plasma concentrations of dexmedetomidine in dogs (Honkavaara et al. 2012) and detomidine and butorphanol in horses (Vainionp a a et al. 2013; Pakkanen et al. 2015). The aims of this equine study were: 1) to examine the clinical effects of romifidine and MK-467, respectively, when administered IV as sole agents; 2) to determine the potential ability of MK-467 to attenuate the cardiovascular and intestinal adverse effects and the level of sedation induced by romifidine when both drugs are administered simultaneously IV; and 3) to analyse plasma concentrations and calculate pharmacokinetic variables of both romifidine and MK-467. Materials and methods The study was approved by the National Animal Experimentation Board of Finland (ESA VI/7608/ 04.10.03.2012). Seven adult, American Society of 600 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610

Anesthesiologists (ASA) class I, non-pregnant Finnhorse mares were included. The sample size was based on practical considerations and not on a formal power calculation because results of previous comparable cross-over studies in horses indicated that a sample size of seven horses would enable the detection of clinically relevant differences between the treatments for HR, ABP, sedation and drug concentration (Vainionp a a et al. 2013; Pakkanen et al. 2015). During the study, mares were placed in stocks. Two 14 gauge, 80 mm catheters (Intraflon; Vygon UK Ltd, UK) were placed in the right jugular vein with a distance of at least 15 cm between them. The proximal catheter was used for blood sampling for plasma drug concentration measurements and venous blood gas analyses, and the distal catheter was used for the injection of test drug(s), after which it was removed immediately. A 20 gauge, 45 mm catheter (BD Arterial Cannula; Becton Dickinson India Pvt. Ltd, India) was placed in a transverse facial artery for direct ABP monitoring and arterial blood gas sampling. All catheters were placed after clipping, surgical preparation and local infiltration of the skin with lidocaine 2% (Lidocain; Orion Corp., Finland). The arterial catheter was connected to an electronic pressure transducer with a zero reference point at the level of the shoulder. Base apex electrocardiogram (ECG) data and ABP were continuously monitored (S/ 5 multi-parameter monitor; Datex Ohmeda Ltd, UK). ECG data were stored for retrospective analysis. All mares were subjected to three treatment protocols. According to a Latin square randomized crossover design with a minimal wash-out period of 6 days, mares were allocated to receive 80 lg kg 1 of romifidine (Sedivet; Boehringer Ingelheim, Vetmedica GmbH, Germany) (R) or 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467 (Vetcare Ltd, Finland) (R + MK) or 200 lg kg 1 of MK-467 (MK). MK-467 was supplied as a powder and mixed to a solution with a final concentration of 10 mg ml 1 with sterile 0.9% saline (Natriumklorid; B Braun Melsungen AG, Germany) on the morning of each study day. All drugs were made up to a total volume of 20 ml by the addition of 0.9% saline and administered IV over 15 seconds using a single syringe. Cardiovascular and respiratory variables At baseline [5 minutes before injection of the test drug(s), T 5 ], and at 3, 5, 10, 20, 30, 40, 50, 60, 70, 80 and 90 minutes after injection (T 3,T 5,T 10,T 20, T 30,T 40,T 50,T 60,T 70,T 80 and T 90, respectively), the following data were recorded: systolic, mean and diastolic ABP (SAP, MAP and DAP, respectively); HR derived from the ECG, and respiratory rate (f R ). Respiratory rate was calculated by counting chest or abdominal excursions over a period of 30 seconds. The recorded ECGs were analysed retrospectively for abnormalities in cardiac conduction for the period during which the mares were confined to the stocks. Arterial and venous blood gas samples (PICO50 blood gas syringes; Radiometer Medical ApS, Denmark) were taken at T 5 and at T 5,T 15,T 30 and T 60 after injection for measurement of oxygen tension (PaO 2 and PvO 2, respectively), carbon dioxide tension (PaCO 2 and PvCO 2, respectively) and ph (ph a and ph v, respectively) (IDEXX VetSta; ME, USA). Sedation and borborygmi scoring Sedation was assessed and auscultation of borborygmi performed by an investigator unaware of treatment allocation. Sedation was scored using a simple descriptive scale after Rohrbach et al. (2009) at T 5,T 1,T 3,T 5,T 10,T 15,T 30,T 45,T 60,T 75 and T 90. After the horses had been returned to their stables, sedation was scored again at T 120,T 180 and T 240. Individual sedation scores were summed to achieve a final sedation score using a scale of 0 10 on which 0 represents no sedation at all and 10 represents heavy sedation but still standing (Table S1). For analysis, a relative sedation score was obtained by subtracting the baseline score from the scores attained at the various time-points. Sedation was defined as a relative sedation score of >3. Following sedation scoring, auscultation of the left and right dorsal and ventral abdomen was performed at T 5,T 10,T 30,T 60,T 90,T 120,T 180 and T 240. After Mama et al. (2009), a numerical score equal to the number of borborygmi auscultated over a 30 second period was assigned to each quadrant; a score of 0.5 was used when uncoordinated rumbling or gaseous sounds were present. A single borborygmi score was achieved by summation of all four individual quadrant scores. Other relevant characteristics occurring during the period for which mares were confined to the stocks, such as defecating, urinating and restlessness, were noted. Plasma drug concentrations To facilitate the determination of romifidine and MK- 467 plasma concentrations, venous blood samples 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610 601

were collected into EDTA tubes at T 10,T 20,T 30,T 60, T 90,T 120,T 180 and T 240. Plasma was separated by centrifugation and stored at 20 C within 2 hours of collection. Concentrations of MK-467 were determined using high-performance liquid chromatography mass spectrometry (HPLC-MS/MS) after solid phase extraction with Sep-Pak tc18 96-well extraction plates (Waters Corp., MA, USA), using RS- 79948 (Tocris Bioscience Ltd, UK) as the internal standard. Reversed-phase separation with a SunFire C18 column (2.1 9 150 mm, 3.5 lm; Waters Corp.) and a gradient solvent system (0.1% formic acid in water and acetonitrile) was used. Quantitative detection was performed in multi-reaction monitoring mode with a triple quadrupole mass spectrometer (API4000; MDS Sciex, ON, Canada). The chromatograms were analysed and processed using Analyst Version 1.4.1 [developed jointly by Applied Biosystems, Inc. (CA, USA) and MDS Sciex]. Concentrations of romifidine were analysed employing the same procedure, but without an internal standard. The reference material used for quantitation was the commercial drug formulation given to the mares. For MK-467 and RS-79948, the respective precursor ions scanned were m/z 419.3 and m/z 365.3. The fragment ions monitored and used for quantitation were m/z 200.1 for MK-467 and m/z 190.2 for RS-79948. The linearity of the assay lay in the range of 10 900 ng ml 1. The acceptance criterion for intra-assay accuracy of back-calculated calibration standards and quality control samples (at three different concentration levels) was 15%. The scanned precursor ion for romifidine was m/z 258.1 and the fragment ion monitored and used for quantitation was m/z 160.1. Peak areas of romifidine and quadratic regression with weighting of 1/x 2 were used for calculations. The range of the assay was 0.88 88 ng ml 1. The acceptance criterion for intra-assay accuracy of back-calculated calibration standards and quality control samples (at three different concentration levels) was 20%. Statistical analysis Categorical data (scores) are reported as the median and range, and continuous data as the mean s- tandard deviation (SD) and were checked for normal distribution using the Shapiro Wilk test. Linear mixed modelling for repeated measures was performed to test whether combined administration of MK-467 and romifidine attenuated the detrimental side effects of romifidine, using SPSS Version 22.0 (IBM Corp., NY, USA). For each clinical variable, fixed effects were analysed to assess mean differences among the three treatments over the entire measurement period. Within each treatment, mean differences were analysed between each time interval and baseline in the same horse. The variables drug and time were considered as fixed effects, time was the repeated measure, and an interaction term for drug and time was added to each model. To adequately model the correlation pattern of the data, effects of different covariance structures were evaluated; autoregressive (order 1) was the best choice of covariance structure and was therefore selected and used in the analyses. In all analyses, restricted maximum likelihood estimates of variables were obtained for the specified covariance structure. Contrasts of least square means for comparisons between drugs, and within treatment for select time intervals, were performed using Bonferroni correction. Areas under the time curves (AUCs) for sedation and borborygmi scores were analysed using one-way analysis of variance (ANOVA) for repeated measures, with Bonferroni s multiple comparison test, using GraphPad Prism for Windows Version 5.01 (GraphPad Software, Inc., CA, USA). Pharmacokinetic data were calculated using WinNonlin Professional Version 5.3 (Pharsight Corp., CA, USA), according to a non-compartmental model. The paired Student s t-test was used for pairwise comparisons of pharmacokinetic parameters between treatments. A two-sided p-value of <0.05 was considered to indicate statistical significance. Results The mean SD age of the mares was 15 4 years; mean SD body weight was 586 40 kg. Cardiovascular and respiratory variables Because of a technical failure, cardiovascular parameters in one horse treated with R could not be measured at T 3 and T 5 ; ECG data for the same horse were also not available for retrospective analysis. In comparison with baseline values, after R, SAP and DAP increased significantly from T 3 to T 50 ; MAP increased significantly from T 5 to T 50. Both HR and f R decreased significantly during the whole measurement period. After R + MK, significant decreases in 602 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610

SAP (T 3 to T 50 and T 70 ), MAP (T 3 to T 40 ), HR and f R (T 3 to T 90 ) occurred; DAP did not change. Administration of MK did not change ABP, but significantly increased both HR (T 3,T 5,T 20 to T 40 ) and f R (T 3,T 5, T 20,T 40 and T 90 ) (Table 1). After R, SAP, MAP and DAP were significantly higher than after R + MK. SAP was higher after MK than after R + MK. DAP was higher after R than after MK. HR and f R were significantly higher after MK than after either R or R + MK. HR was lower after R than after R + MK, although this difference was not significant (Table 1). In all horses, second-degree atrioventricular (AV) blocks occurred within 1 minute after R and lasted up to 54 minutes; in some mares a delay in cardiac conduction and absence of P waves with presence of normal QRS configuration were noticed. In all but one horse, second-degree AV blocks were noted after R + MK; these disappeared within 3 minutes. No changes in cardiac conduction occurred after MK. Blood gases Significant changes over time were seen in PaO 2, PvO 2, PaCO 2 and PvCO 2. After R + MK, PaCO 2 was significantly higher than after MK (p = 0.009). PvO 2 was lower after R than after R + MK (p < 0.001) and MK (p = 0.000). PvCO 2 was lower after MK than after R (p = 0.001) and R + MK (p = 0.009), and ph v was higher after MK than after R (p = 0.013) and R + MK (p = 0.009) (Table 2). Sedation scores Maximum median (range) sedation scores corrected for baseline were 7 (5 8) after R (T 15 ), 5 (3 7) after R + MK (T 5 to T 30 ) and 1 (0 1) after MK (T 120 to T 240 ). Compared with baseline, significant increases over time were found for both R at T 1 to T 75 (p < 0.0001) and T 90 (p = 0.002), and for R + MK at T 1 (p = 0.000), T 3 to T 45 (p < 0.0001), T 60 (p = 0.006) and T 90 (p = 0.031) (Fig. 1). The AUCs for sedation time were significantly higher (p < 0.05) after R (503) and R + MK (379) than after MK (109); no significant difference was detected between R and R + MK. Borborygmi scores Intestinal activity decreased significantly for 180 minutes after R (T 10 to T 120, p < 0.0001; T 180, p = 0.022). Significant reductions were also noticed at T 10 (p = 0.022), T 30 (p = 0.014) and T 90 (p = 0.008) following R + MK, although to a lesser extent than after R. No significant changes over time were detected after MK (Fig. 2). The AUC for borborygmi time was significantly smaller (p < 0.05) after R (500) than after MK (1428) and R + MK (968); no significant difference was detected between R + MK and MK. All horses defecated noticeably more often after MK than after the other two treatments. After MK, three horses displayed signs of restlessness and kicked their pelvic limbs towards the abdomen; two of these horses had watery faeces. Pharmacokinetic data Plasma drug concentrations are shown in Fig. 3. The plasma concentration of MK-467 in one horse at T 240 has been excluded because of an exceptionally high reading, possibly resulting from contamination or a technical error. Elimination half-life (T ½ ), volume of distribution (V z ) and clearance of romifidine were significantly greater and the AUC for concentration extrapolated to infinity (AUC inf ) was significantly smaller after R + MK than after R (Table 3). No significant differences between R + MK and MK were found for the pharmacokinetic variables calculated for MK-467. Discussion Intravenous romifidine at a dose of 80 lg kg 1 resulted in a characteristic increase in ABP and decrease in HR as have been reported previously in the same species (Gasthuys et al. 1990; Clarke et al. 1991; England et al. 1992; Freeman & England 2000; Freeman et al. 2002; Wojtasiak-Wypart et al. 2012). However, the biphasic effect of romifidine on ABP as documented by Clarke et al. (1991) and Wojtasiak-Wypart et al. (2012) was not observed; at the end of the measurement period ABP was more or less comparable with baseline values. Sympathetic outflow is decreased by a 2 -agonists by both activation of a 2A adrenoceptors located in the locus coeruleus in the central nervous system, and presynaptic inhibition of noradrenaline release from peripheral sympathetic nerve endings by activation of a 2B and a 2C adrenoceptors (autoreceptors) (McCallum et al. 1998; Kamibayashi & Maze 2000). The decrease in sympathetic tone, probably combined with inhibition of ganglionic transmission, can result in (relative) hypotension (McCallum et al. 1998). The various a 2 -agonists 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610 603

Table 1 Cardiopulmonary variables (mean SD) in seven adult Finnhorse mares after intravenous romifidine (R), romifidine and MK-467 (R + MK), and MK-467 (MK) at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467 SAP (mmhg) MAP (mmhg) DAP (mmhg) HR (beats minute 1 ) fr (breaths minute 1 ) Variable Treatment R R + MK, MK R R + MK MK R, R + MK MK R R + MK MK, R R + MK MK, T0 157 18 163 17 166 14 121 8 123 10 124 14 96 7 98 9 98 12 48 2 45 8 50 10 33 10 27 14 22 13 T3 179 31* 141 22* 155 14 123 24 108 19* 116 12 108 16* 90 16 88 10 30 8* 31 4* 64 14* 13 5* 10 4* 31 11* (p = 0.003) (p = 0.004) (p = 0.008) (p = 0.037) (p < 0.0001) (p = 0.000) (p = 0.000) (p = 0.000) (p = 0.001) (p = 0.049) T5 189 35* 143 24* 167 8 141 23* 109 18* 123 11 118 19* 91 13 94 10 25 6* 32 4* 60 12* 14 6* 13 4* 31 16* (p < 0.0001) (p = 0.006) (p = 0.000) (p = 0.009) (p = 0.000) (p < 0.0001) (p = 0.001) (p = 0.014) (p = 0.002) (p = 0.003) (p = 0.049) T10 191 31* 138 29* 160 9 136 15* 108 20* 119 8 115 11* 90 18 94 4 26 6* 31 4* 57 13 18 9* 11 6* 27 9 (p < 0.0001) (p = 0.001) (p = 0.009) (p = 0.008) (p = 0.001) (p < 0.0001) (p = 0.000) (p = 0.015) (p = 0.001) T20 177 26* 138 29* 175 11 136 15* 108 18* 125 9 116 11* 88 14 92 8 26 5* 32 5* 62 15* 14 6* 9 3* 34 12* (p = 0.009) (p = 0.001) (p = 0.009) (p = 0.005) (p = 0.001) (p < 0.0001) (p = 0.001) (p = 0.001) (p = 0.001) (p = 0.000) ( (p = 0.013) T30 173 25* 138 25* 164 16 137 15* 107 19* 121 7 115 11* 89 14 96 11 28 4* 33 5* 68 19* 11 1* 7 2* 27 9 (p = 0.037) (p = 0.001) (p = 0.004) (p = 0.003) (p = 0.001) (p < 0.0001) (p = 0.002) (p < 0.0001) (p < 0.0001) (p < 0.0001) T40 180 17* 141 29* 159 15 143 14* 110 21* 121 13 121 19* 91 16 95 16 29 3* 34 4* 60 18* 14 6* 6 1* 33 16* (p = 0.003) (p = 0.003) (p < 0.0001) (p = 0.019) (p < 0.0001) (p < 0.0001) (p = 0.002) (p = 0.014) (p = 0.001) (p < 0.0001) (p = 0.024) T50 173 22* 147 19* 169 9 141 12* 114 15 128 10 121 10* 92 14 103 10 30 4* 34 5* 56 10 10 3* 7 2* 23 5 (p = 0.033) (p = 0.035) (p = 0.001) (p = 0.000) (p < 0.0001) (p < 0.0001) (p < 0.0001) (p < 0.0001) T 60 163 4 150 22 166 10 127 8 114 16 123 8 104 9 92 15 96 6 31 3* 33 4* 51 12 10 3* 6 1* 29 15 (p < 0.0001) (p = 0.002) (p < 0.0001) (p < 0.0001) T70 163 11 148 23* 170 11 126 8 114 18 129 8 100 11 96 13 103 5 33 7* 34 6* 48 10 8 2* 6 1* 29 15 (p = 0.040) (p = 0.000) (p = 0.000) (p < 0.0001) (p < 0.0001) T 80 159 10 157 23 168 10 121 12 120 19 127 9 99 13 99 12 101 8 34 4* 35 4* 54 12 9 3* 7 2* 24 10 (p = 0.001) (p = 0.004) (p < 0.0001) (p < 0.0001) T90 157 7 159 22 169 12 123 8 119 18 124 8 102 9 95 10 97 9 32 5* 37 9* 50 8 10 4* 8 2* 37 16* (p < 0.0001) (p < 0.0001) (p < 0.0001) (p < 0.0001) (p = 0.002) SAP, systolic arterial blood pressure; MAP, mean arterial blood pressure; DAP, diastolic arterial blood pressure; HR, heart rate; fr, respiratory rate. *Statistically significant difference from baseline value within treatment. Statistically significant difference between R + MK and MK. Statistically significant difference between R and R + MK. Statistically significant difference between R and MK. 604 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610

Table 2 Arterial and venous blood gas analyses (mean standard deviation) in seven adult Finnhorse mares after intravenous romifidine (R), romifidine and MK-467 (R + MK), and MK-467 (MK) at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467. Given p-values are for changes over time within each individual treatment; see text for p-values between treatments Parameter Treatment Baseline T 5 T 15 T 30 T 60 PaO 2 (mmhg) R 105 8 93 13* (p = 0.008) 92 6* (p = 0.004) 97 5 98 4 R + MK 108 9 100 8* (p = 0.05) 100 8* (p = 0.05) 101 6 102 6 MK 103 4 102 4 98 2 98 4 95 4 PvO 2 (mmhg) R, 40 5 30 2* (p < 0.0001) 34 2* (p = 0.002) 37 1 41 2 R + MK 39 3 42 4 44 3* (p = 0.012) 45 4* 42 3 (p = 0.003) MK 39 1 46 3* (p = 0.000) 45 4* (p = 0.003) 42 3 42 5 PaCO 2 (mmhg) R 43 2 40 3 43 2 44 3 46 3* (p = 0.019) R + MK 41 2 44 2 45 1* (p = 0.015) 47 2* (p = 0.000) 47 3* (p = 0.000) MK 42 1 42 2 43 1 42 1 44 2 PvCO 2 (mmhg) R 45 3 48 2* (p = 0.001) 49 2* (p < 0.0001) 49 2* (p < 0.0001) R + MK 45 2 47 1 47 2* (p = 0.035) 48 2* (p = 0.004) 49 2* (p < 0.0001) 48 2* (p = 0.001) MK, 45 2 43 1* (p = 0.015) 44 1 44 2 44 1 ph a R 7.44 0.01 7.45 0.02 7.44 0.02 7.44 0.02 7.45 0.02 R + MK 7.46 0.02 7.44 0.01 7.43 0.02 7.43 0.01 7.44 0.02 MK 7.45 0.01 7.45 0.01 7.45 0.01 7.46 0.01 7.46 0.01 ph v R 7.44 0.02 7.43 0.02 7.43 0.01 7.43 0.01 7.44 0.01 R + MK 7.44 0.02 7.43 0.01 7.43 0.02 7.43 0.01 7.44 0.01 MK, 7.45 0.01 7.45 0.01 7.45 0.01 7.46 0.01 7.46 0.01 PaO 2, arterial oxygen tension; PvO 2, venous oxygen tension; PaCO 2, arterial carbon dioxide tension; PvCO 2, venous carbon dioxide tension. *Statistically significant difference from baseline value within treatment. Statistically significant difference between R + MK and MK. Statistically significant difference between R and R + MK. Statistically significant difference between R and MK. differ in their a 2 :a 1 -adrenoceptor selectivity ratio. It is likely that romifidine is not very selective and only acts as a partial a 2 -agonist (Wojtasiak-Wypart et al. 2012); its activation of a 1 -receptors may have contributed to the absence of a hypotensive phase. Atrioventricular conduction disturbances such as second-degree AV block occurred almost instantaneously after romifidine and lasted for almost an hour in some of the horses, which is comparable with findings documented by others (Gasthuys et al. 1990; Figueiredo et al. 2005; Wojtasiak-Wypart et al. 2012). Freeman et al. (2002) and Wojtasiak-Wypart et al. (2012) reported a significant decrease in CO (up to 56%) after romifidine, lasting for at least 30 and 90 minutes, respectively. This decrease in CO is most likely the result of an increase in SVR, inducing a baroreceptor-mediated bradycardia, and may potentially result in markedly reduced peripheral perfusion. The combined administration of MK-467 with romifidine attenuated the romifidine-induced bradycardia and increase in ABP. These results are comparable with those reported by Vainionp a a et al. (2013), who found that MK-467 prevented detomidine-induced bradycardia in horses, and with those in other species in which the peripheral antagonist was administered with medetomidine or dexmedetomidine (Bryant et al. 1998; Enouri et al. 2008; Honkavaara et al. 2008, 2011; Raekallio et al. 2010; Rolfe et al. 2012; Salla et al. 2014). In the study by Vainionp a a et al. (2013), MK-467 prevented the detomidine-induced cardiac conduction disturbances as early as 1 minute after their combined administration. In the current study, the concomitant administration of MK-467 resulted in a clear reduction in the duration of romidifine-induced conduction disturbances, but did not completely abolish them. At the doses used, differences in early drug distribution may have played a role; romifidine may have had a greater peripheral effect than MK- 467 in the immediate post-administration period. Because the effects of MK-467 are dose-dependent, a higher dose of MK-467 may have completely 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610 605

Figure 1 Median sedation scores in seven adult Finnhorse mares after intravenous romifidine (R), romifidine and MK-467 (R + MK), and MK-467 (MK) at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467. *Significant change over time compared with baseline within group R. Significant change over time compared with baseline within group R + MK. Figure 2 Median borborygmi scores in seven adult Finnhorse mares after intravenous romifidine (R), romifidine and MK-467 (R + MK), and MK-467 (MK) at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467. *Significant change over time compared with baseline within group R. Significant change over time compared with baseline within group R + MK. prevented romifidine-induced bradycardia and conduction disturbances. Honkavaara et al. (2011) came to a similar conclusion when dexmedetomidine was combined with various doses of MK-467 in dogs. Both HR and ABP were relatively high at baseline in all horses in the current study, most likely because of a stress response. The significant decrease in HR compared with baseline after R + MK may be largely explained by the centrally mediated sedative and sympatholytic effects of romifidine resulting in HR returning to normal resting values. Correspondingly, the decrease in SAP and MAP from baseline in horses in the R + MK treatment was probably the result of the romifidine-induced decrease in HR. Administration of MK-467 as the sole agent resulted in behavioural stimulation, evidenced by an increase in HR and f R and accompanied by restlessness in some horses. The cause of this stimulation is not known with certainty, but increased sympatho-adrenal activity mediated by the removal of presynaptic autoreceptor inhibition, may have been involved. Centrally mediated effects are an unlikely cause because of the drug s low lipid solubility. DiMaio Knych et al. (2011) reported a significant increase in HR in horses after IV administration of yohimbine, a centrally acting a 2 - antagonist. These authors did not find a correlation between episodes of increased HR and behavioural excitation, and suggested that the increase in HR may have been the result of a peripheral effect on a 2 - adrenoceptors or perhaps another receptor type, increasing noradrenaline release and therefore sympathetic tone. The same theory may be applicable to MK-467. Respiratory rate decreased after both R and R + MK, probably as a result of the sedative effect of romifidine. Alternatively, this decrease may have resulted from a decrease in ventilatory responsiveness or a decrease in receptor sensitivity leading to a 606 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610

significant decrease in PaO 2 after both treatments. The increase in PaCO 2 and decrease in PaO 2 were probably the result of both hypoventilation and (a) (b) Figure 3 Plasma concentrations (mean standard deviation) in seven adult Finnhorse mares of (a) romifidine after intravenous (IV) administrations of romifidine alone (R) and in combination with MK-467 (R + MK), and (b) MK- 467 after IV administrations of MK-467 alone (MK) and as R + MK, all at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467. increased ventilation/perfusion mismatching within the lungs, as was reported by Nyman et al. (2009), who found similar changes after detomidine administration in horses. However, the changes in PaO 2 and PaCO 2 in the current study are unlikely to be clinically relevant as all values remained within or close to normal reference ranges (Soma et al. 1996; Aguilera-Tejero et al. 1998). After R, PvO 2 was lower than after the other treatments, in which it was always higher than baseline. Freeman et al. (2002) reported a significant decrease in mixed venous oxygen partial pressure after romifidine at the same dose as used in the present study. Mixed venous oxygen partial pressure can be used to assess tissue oxygen delivery and blood flow, and depends on arterial oxygen concentration, CO and SVR. Although the venous blood gases in the current study may not be representative of mixed venous values, the better cardiovascular function after R + MK compared with after R may have prevented romifidine-induced reduction of peripheral perfusion and oxygen delivery and a consequently increased oxygen extraction ratio as was suggested by Nyman et al. (2009). Although all horses were sedated in the present study for at least 90 minutes after both R and R + MK, administration of the latter resulted in slightly lower median sedation scores. This reduced sedative effect could be explained by the altered pharmacokinetics of romifidine as a result of improved cardiovascular function: a higher CO and better tissue perfusion resulted in enhanced drug distribution, increased clearance and hence a lower plasma concentration and therefore sedative effect. The effect of MK-467 on the quality of sedation induced by other a 2 -agonists has been reported in other publications. Vainionp a a et al. (2013) investigated the effects of the combined IV administration Table 3 Pharmacokinetic variables (mean standard deviation) of romifidine when administered intravenously (IV) alone (R) and in combination with MK-467 (R + MK), and of MK-467 when administered IV alone (MK) and in combination with romifidine (R + MK) at dosages of 80 lg kg 1 of romifidine and 200 lg kg 1 of MK-467 in seven adult Finnhorse mares Treatment T ½ (minutes) AUC inf (minutes ng ml 1 ) V Z (ml kg 1 ) Clearance (ml minutes 1 kg 1 ) Romifidine R 75.6 3.0* 3772 421* 2333 224* 21.4 2.4* R + MK 89.1 12.8* 2977 285* 3473 488* 27.1 2.7* MK-467 MK 141 28.6 34,157 5652 1189 121 6.0 0.99 R + MK 170 82.7 39,003 12,253 1217 188 5.5 1.6 T ½, elimination half life; AUC inf, area under the concentration-time curve extrapolated to infinity; V z, volume of distribution. *Statistically significant difference between R and R + MK. 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610 607

of MK-467 and detomidine in horses and found it provided a clinical quality of sedation similar to that facilitated by detomidine alone, albeit with a slightly shorter duration of effect. In dogs, the level of clinical sedation was slightly reduced by MK-467 in combination with dexmedetomidine (Restitutti et al. 2011). Honkavaara et al. (2008) did not find significant differences in the AUCs for sedation between IV dexmedetomidine and dexmedetomidine combined with MK-467. When romifidine is combined with MK-467 in horses, a higher dose of the former may be needed to achieve a level of sedation similar to that achieved by romifidine alone. The effect of romifidine on intestinal activity was very discernible: the borborygmi score was significantly reduced during the whole monitoring period. Although a decrease in gastrointestinal motility may be desirable in some conditions, such as spastic abdominal disorders, depressant effects should ideally be avoided in horses already suffering from intestinal hypomotility or in those at risk for developing postoperative ileus. The combined administration of R and MK-467 also resulted in reduced borborygmi, but to a much lesser extent and for a much shorter time than romifidine alone. This may indicate that MK-467 only partially prevented the effect of romifidine on gut motility; a different ratio of romifidine to MK-467 may perhaps totally abolish the effect of romifidine on the occurrence of borborygmi. No statistically significant changes in borborygmi over time occurred after MK, but increased defecation with relatively watery faeces suggests that MK did affect intestinal motility. Zullian et al. (2011) assessed the effects of the a 2 -antagonists atipamezole and yohimbine on equine small intestinal motility. Neither antagonist modified either spontaneous or electrically induced jejunum contractions in vitro, indicating that they have no intrinsic activity on a 2 -receptors located within the equine gastrointestinal tract. By contrast, DiMaio Knych et al. (2011) reported an increase in borborygmi and bowel movements after IV administration of yohimbine. In line with the current results, dexmedetomidine was found to markedly inhibit gastric emptying and oro caecal transit in healthy human volunteers (Iirola et al. 2011). Further, a 2 -antagonists such as yohimbine have been shown to enhance acetylcholine release in humans, thereby increasing colonic tone and inducing colonic contractions (Bharucha et al. 1997). This may explain the increased incidence of defecation seen in all horses treated with MK-467, as well as the signs of abdominal discomfort noted in some of them. Greater gut motility may be another explanation for the increased HR seen after MK. Additional studies are needed to elucidate the mechanisms responsible for the increased HR and the effect on gut motility of MK-467. The haemodynamic effects of a 2 -agonists affect their own disposition, especially immediately after IV administration (Salonen et al. 1995; Dutta et al. 2000; Honkavaara et al. 2012; Vainionp a a et al. 2013). This suggests that their pharmacokinetic profile is non-linear and therefore a non-compartmental model was used in the current study, although Wojtasiak-Wypart et al. (2012) suggested a two-compartmental model as the best fit for IV romifidine administration. The concurrent administration of MK-467 increased T ½ and V z, decreased AUC inf and accelerated romifidine clearance. These changes in the disposition of romifidine are in agreement with those in previous publications in which MK-467 was reported to alter the pharmacokinetic behaviour of other a 2 -adrenoceptor agonists, such as detomidine in horses (Vainionp a a et al. 2013; Pakkanen et al. 2015), and dexmedetomidine in dogs (Honkavaara et al. 2012). The improved cardiac function and peripheral blood flow induced by MK-467 (Restitutti et al. 2013) result in faster and more efficient distribution of the a 2 -agonist into tissues; clearance is likely to be enhanced by increased hepatic circulation and drug metabolism (Salonen et al. 1995). The pharmacokinetic variables of MK-467 when administered IV as the sole agent have been reported only in dogs (Honkavaara et al. 2012). In the current study, MK-467 reversed the romifidineinduced cardiovascular depression; therefore romifidine did not significantly affect the disposition of MK-467 and no differences in its pharmacokinetics were detected between R + MK and MK. Conclusions The combined administration of romifidine and MK-467 in horses prevented the characteristic cardiovascular changes commonly seen with a 2 -adrenoceptor agonists, but did not affect the quality of sedation. MK-467 also attenuated the typical romifidine-induced decrease in intestinal motility. These results are potentially desirable in horses with clinical disease or when used as part of a general anaesthetic protocol, in which a reduction in CO, peripheral perfusion or intestinal activity may be 608 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610

detrimental. However, further study is warranted to determine the optimal dose ratio between romifidine and MK-467 in horses. Acknowledgements Vetcare Ltd, Finland supplied the MK-467 and funded some of the materials used in this study. Romifidine was donated by Boehringer Ingelheim, UK. The blood gas analyser was provided on loan by IDEXX, Finland. The authors would like to thank Marianna Myllym aki, MTT Agrifood Research, Finland, for assistance with the horses, and Lauri Vuorilehto, University of Turku, Finland, for developing the assay methods and performing the concentration analyses. Authors contributions AdV contributed to the study design, the collection, analysis and interpretation of data, and the writing of the paper. SAEP contributed to the study design, and the collection, analysis and interpretation of data. MRR and OMV contributed to the study design and supervised the study. AE contributed to the statistical analysis. MS and PMT contributed to the interpretation of data. All authors contributed to the critical revision of the paper and approved the final manuscript for publication. The authors have no conflicts of interest to report. References Aguilera-Tejero E, Estepa JC, Lopez I et al. (1998) Arterial blood gases and acid-base balance in healthy young and aged horses. Equine Vet J 30, 352 354. Bharucha AE, Camilleri M, Zinsmeister AR et al. (1997) Adrenergic modulation of human colonic motor and sensory function. Am J Physiol 27, 997 1006. Bryant CE, Thompson J, Clarke KW (1998) Characterisation of the cardiovascular pharmacology of medetomidine in the horse and sheep. Res Vet Sci 65, 149 154. Clarke KW, England GCW, Goossens L (1991) Sedative and cardiovascular effects of romifidine, alone and in combination with butorphanol, in the horse. J Vet Anaesth 18, 25 29. Clineschmidt BV, Pettibone DJ, Lotti VJ et al. (1988) A peripherally acting alpha-2 adrenoceptor antagonist: L- 659,066. J Pharmacol Exp Ther 245, 32 40. DiMaio Knych HK, Steffey EP, Stanley SD (2011) Pharmacokinetics and pharmacodynamics of three intravenous doses of yohimbine in the horse. J Vet Pharmacol Ther 34, 359 366. Dutta S, Lal R, Karol MD et al. (2000) Influence of cardiac output on dexmedetomidine pharmacokinetics. J Pharm Sci 89, 519 527. Elfenbein JR, Sanchez LC, Robertson SA et al. (2009) Effect of detomidine on visceral and somatic nociception and duodenal motility in conscious adult horses. Vet Anaesth Analg 36, 162 172. England GCW, Clarke KW, Goossens L (1992) A comparison of the sedative effects of three a 2 -adrenoceptor agonists (romifidine, detomidine and xylazine) in the horse. J Vet Pharmacol Ther 15, 194 201. Enouri SS, Kerr CL, McDonell WN et al. (2008) The effects of a peripheral a 2 adrenergic-receptor antagonist on the hemodynamic changes induced by medetomidine administration in conscious dogs. Am J Vet Res 69, 728 736. Figueiredo JP, Muir WW, Smith J et al. (2005) Sedative and analgesic effects of romifidine in horses. Int J Appl Res Vet Med 3, 249 258. Freeman SL, England GCW (2000) Investigation of romifidine and detomidine for the clinical sedation of horses. Vet Rec 147, 507 511. Freeman SL, England GCW (2001) Effect of romifidine on gastrointestinal motility, assessed by transrectal ultrasonography. Equine Vet J 33, 570 576. Freeman SL, Bowen IM, Bettschart-Wolfensberger R et al. (2002) Cardiovascular effects of romifidine in the standing horse. Res Vet Sci 72, 123 129. Gasthuys F, Parmentier D, Goossens L et al. (1990) A preliminary study on the effects of atropine sulphate on bradycardia and heart blocks during romifidine sedation in the horse. Vet Res Commun 14, 489 502. Honkavaara J, Raekallio M, Kuusela E et al. (2008) The effects of L-659,066, a peripheral a 2 -adrenoceptor antagonist, on dexmedetomidine-induced sedation in dogs. Vet Anaesth Analg 35, 409 411. Honkavaara JM, Restitutti F, Raekallio MR et al. (2011) The effects of increasing doses of MK-467, a peripheral alpha 2 -adrenergic receptor antagonist, on the cardiopulmonary effects of intravenous dexmedetomidine in conscious dogs. J Vet Pharmacol Ther 34, 332 337. Honkavaara JM, Restitutti F, Raekallio M et al. (2012) Influence of MK-467, a peripherally acting a 2 - adrenoceptor antagonist on the disposition of intravenous dexmedetomidine in dogs. Drug Metab Dispos 40, 445 449. Iirola T, Vilo S, Aantaa R et al. (2011) Dexmedetomidine inhibits gastric emptying and oro-caecal transit in healthy volunteers. Br J Anaesth 106, 522 527. Kamibayashi T, Maze M (2000) Clinical uses of a 2 - adrenergic agonists. Anesthesiology 93, 1345 1349. Mama KR, Grimsrud K, Snell T et al. (2009) Plasma concentrations, behavioural and physiological effects following intravenous and intramuscular detomidine in horses. Equine Vet J 41, 772 777. 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610 609

McCallum JB, Boban N, Hogan Q et al. (1998) The mechanism of a 2 -adrenergic inhibition of sympathetic ganglionic transmission. Anesth Analg 87, 503 510. Nyman G, Marntell S, Edner A et al. (2009) Effect of sedation with detomidine and butorphanol on pulmonary gas exchange in the horse. Acta Vet Scand 51, 22 30. Pakkanen SAE, Raekallio MR, Mykk anen AK et al. (2015) Detomidine and the combination of detomidine and MK- 467, a peripheral alpha-2 adrenoceptor antagonist, as premedication in horses anaesthetized with isoflurane. Vet Anaesth Analg 42, 527 536. Raekallio MR, Honkavaara JM, Vainio OM (2010) The effects of L-659,066, a peripheral a 2 -adrenoceptor antagonist, and verapamil on the cardiovascular influences of dexmedetomidine in conscious sheep. J Vet Pharmacol Ther 33, 434 438. Restitutti F, Honkavaara JM, Raekallio MR et al. (2011) Effects of different doses of L-659,066 on the bispectral index and clinical sedation in dogs treated with dexmedetomidine. Vet Anaesth Analg 38, 415 422. Restitutti F, Laitinen MR, Raekallio MR et al. (2013) Effect of MK-467 on organ blood flow parameters detected by contrast-enhanced ultrasound in dogs treated with dexmedetomidine. Vet Anaesth Analg 40, e48 e56. Roger T, Ruckebusch Y (1987) Colonic alpha 2- adrenoceptor-mediated responses in the pony. J Vet Pharmacol Ther 10, 310 318. Rohrbach H, Korpivaara T, Schatzmann U et al. (2009) Comparison of the effects of the alpha-2 agonists detomidine, romifidine and xylazine on nociceptive withdrawal reflex and temporal summation in horses. Vet Anaesth Analg 36, 384 395. Rolfe NG, Kerr CL, McDonell WN (2012) Cardiopulmonary and sedative effects of the peripheral a 2 -adrenoceptor antagonist MK 0467 administered intravenously or intramuscularly concurrently with medetomidine in dogs. Am J Vet Res 73, 587 594. Salla K, Restitutti F, Vainionp a a M et al. (2014) The cardiopulmonary effects of a peripheral alpha-2- adrenoceptor antagonist, MK-467, in dogs sedated with a combination of medetomidine and butorphanol. Vet Anaesth Analg 41, 567 574. Salonen S, Vuorilehto L, Vainio O et al. (1995) Atipamezole increases medetomidine clearance in the dog: an agonist antagonist interaction. J Vet Pharmacol Ther 18, 328 332. Sciberras DG, Reed JW, Elliott C et al. (1994) The effects of a peripherally selective a 2 -adrenoceptor antagonist, MK- 467, on the metabolic and cardiovascular response to exercise in healthy man. Br J Clin Pharmacol 37, 39 44. Soma LR, Uboh CE, Nann L (1996) Prerace venous acidbase values in Standardbred horses. Equine Vet J 28, 390 396. Vainionp a a MH, Raekallio MR, Pakkanen SAE et al. (2013) Plasma drug concentrations and clinical effects of a peripheral alpha-2-adrenoceptor antagonist, MK- 467, in horses sedated with detomidine. Vet Anaesth Analg 40, 257 264. Wojtasiak-Wypart M, Soma LR, Rupdy JA et al. (2012) Pharmacokinetic profile and pharmacodynamic effects of romifidine hydrochloride in the horse. J Vet Pharmacol Ther 35, 478 488. Zullian C, Menozzi A, Pozzoli C et al. (2011) Effects of a 2 - adrenergic drugs on small intestinal motility in the horse: an in vitro study. Vet J 187, 342 346. Received 21 September 2015; accepted 24 November 2015. Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. System for scoring sedation after Rohrbach et al. (2009). 610 2016 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesia and Analgesia, 43, 599 610