Veterinary Anaesthesia and Analgesia 2017, 44, 86e97 http://dx.doi.org/10.1111/vaa.

12410

RESEARCH PAPER

Pharmacokinetics and pharmacodynamics of

intravenous romifidine and propranolol
administered alone or in combination for
equine sedation

Alessia Cenani*, Robert J Brosnany, Shara Madiganz, Heather K Knychx &

John E Madigan¶
*Veterinary Medical Teaching Hospital, University of California Davis, Davis, CA, USA
yDepartment of Surgical and Radiological Sciences, School of Veterinary Medicine, University of
California Davis, Davis, CA, USA
zDepartment of Animal Science, University of California Davis, Davis, CA, USA
xKL Maddy Equine Analytical Chemistry Laboratory, University of California Davis, Davis, CA, USA
¶Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California
Davis, Davis, CA, USA

Correspondence: Robert J Brosnan, Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of
California Davis, 2112 Tupper Hall, One Shields Avenue, Davis, CA 95616, USA. E-mail: rjbrosnan@ucdavis.edu

Abstract compared with P and R. Both PR and R decreased

behavioral responsiveness and resulted in sedation
Objective Propranolol has been suggested for anx-
for up to 240 and 480 minutes, respectively.
iolysis in horses, but its sedation efficacy and side
Sedation was deeper in PR for the first 16 minutes.
effects, both when administered alone and in com-
Heart rate significantly decreased in all treatments
bination with a2-adrenoceptor agonists, remain
for at least 60 minutes, and PR significantly
undetermined. This study aimed to document the
increased the incidence of severe bradycardia (<20
pharmacokinetics and pharmacodynamics of pro-
beats minute1).
pranolol, romifidine and their combination.
Conclusions and clinical relevance Although not
Study design Randomized, crossover study.
associated with reduced behavioral responsiveness
Animals Six adult horses weighing 561 ± 48 kg. or sedation alone, propranolol augmented romifi-
dine sedation, probably through alterations in
Methods Propranolol (1 mg kg1; treatment P),
romifidine pharmacokinetics, in horses adminis-
romifidine (0.1 mg kg1; treatment R) or their
tered PR. The occurrence of severe bradycardia
combination (treatment PR) were administered
warrants caution in the co-administration of these
intravenously with a minimum of 1 week between
drugs at the doses studied.
treatments. Alertness, behavioral responsiveness
(visual and tactile) and physiologic variables were
measured before and up to 960 minutes after drug Keywords bradycardia, equine, propranolol.
administration. Blood was collected for blood gas romifidine, sedation.
and acid-base analyses and measurement of
Introduction
plasma drug concentrations. Data were analyzed
using repeated-measures analysis of variance or Alpha2-adrenergic receptor agonists (a2-agonists),
Friedman with HolmeSidak and Wilcoxon rank- such as romifidine, are the most widely used periop-
sum tests (p < 0.05). erative class of sedatives in equine practice and are
used either alone or in combination with other drugs.
Results Systemic clearance significantly decreased
Romifidine in horses produces sedation, analgesia
and the area under the concentration-time curve
and muscle relaxation with less ataxia than other a2-
significantly increased for both drugs in PR

86
Romifidine and propranolol PK–PD in horses A Cenani et al.
agonists (England et al. 1992; Moens et al. 2003),
and dose-dependently improves the quality of recov-
ery from isoflurane anesthesia compared with xyla-
zine (Woodhouse et al. 2013). Sedation is mediated
by downregulation of central norepinephrine
following activation of a2-adrenergic receptors in the
locus coeruleus (Doze et al. 1989). However, a2-
adrenoceptor-mediated sympatholysis (Wang et al.
1994), increased systemic vascular resistance with
reflex increased vagal tone (Gasthuys et al. 1990),
and direct inhibition of hyperpolarization-activated
cyclic nucleotide-gated (HCN) channels (Knaus
et al. 2007) within the sinoatrial node together can
produce profound and persistent bradycardia,
hypertension and reduced cardiac output (CO) in
horses (Wojtasiak-Wypart et al. 2012). Ataxia and
gastrointestinal stasis are also common side effects of
a2-agonists (Freeman & England 2001).
The co-administration of two or more sedative
and/or anxiolytic drugs at lower doses often reduces
the dose-dependent side effects of each drug. Pro-
pranolol is a b1- and b2-adrenoceptor blocker that
attenuates adrenergic activity throughout the body.
In humans, propranolol reduces amygdala hyper-
activation to stressful stimuli (Hurlemann et al.
2010) in patients with anxiety disorders (Etkin &
Wager 2007); anxiolysis is likely to be attributable
to the blockade of locus coeruleus norepinephrine
input to the amygdala (Hurlemann et al. 2010).
Propranolol also modulates the amygdalae
hippocampus interactions during declarative mem-
ory encoding and consolidation of stimuli (Strange &
Dolan 2004). In horses, propranolol has been used to
control supraventricular tachyarrhythmia (Muir &
McGuirk 1985). Bradycardia and decreased cardiac
contractility are known adverse effects of this drug
(Muir & McGuirk 1985). Propranolol has been pro-
posed anecdotally as a potentially useful equine
anxiolytic, tranquilizer or sedative; however, the
utility of propranolol for these purposes, either alone
or in combination with other a2-adrenoceptor ago-
nists, has not been studied experimentally.
The aim of this study was to describe the phar-
macokinetic and pharmacodynamic profile of romi-
fidine, propranolol and their co-administration
following a single intravenous (IV) bolus injection in
awake horses. We hypothesized that propranolol
would both cause sedation and enhance romifidine
sedation. Because each drug can cause bradyar-
rhythmias, we also hypothesized that co-
administration of propranolol and romifidine would
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
exacerbate this adverse effect compared with the
administration of either drug alone.
Materials and methods
Animals
Six horses consisting of four mares and two geldings
(three Thoroughbreds, two Quarter Horses, one
Dutch Warmblood) with a mean ± standard devia-
tion (SD) age of 10 ± 5 years and weight of 561 ± 48
kg were studied in a randomized crossover trial
approved by the Institutional Animal Care and Use
Committee at the University of California Davis, CA,
USA. All horses were deemed healthy based on his-
tory and physical examination.
Horses were randomly assigned to the adminis-
tration of propranolol IV (1 mg kg1; treatment P),
romifidine IV (0.1 mg kg1; Sedivet 1%; Boehringer
Ingelheim VetMedica, Inc., MO, USA) (treatment R),
or IV propranolol (1 mg kg1) and romifidine (0.1 mg
kg1) (treatment PR), with an interval of at least 1
week between treatments. Propranolol is commer-
cially available only as a 1 mg mL1 injectable so-
lution and therefore non-pharmaceutical grade
propranolol was prepared as a 25 mg mL1 solution
as follows: to a quantity of (±) propranolol hydro-
chloride powder of 99% purity (Sigma-Aldrich
Corp., MO, USA) was added 19.75 times the mass of
sterile, distilled and deionized water plus 0.25 times
the mass of a sterile 50% solution of citric acid
(Sigma-Aldrich Corp.). These were mixed in a sterile
beaker until the initial white slurry became clear.
Finally, additional sterile water equal to 20 times the
mass of the initial propranolol was added and mixed,
yielding a final solution of 25 mg mL1 propranolol,
which was passed through a 0.2 mm filter (Millex-GS;
Millipore Corp., MA, USA) and stored in a sterile
multi-dose vial (Allergy Laboratories, Inc., OK, USA)
for up to 24 hours before use. Romifidine was diluted
from its original packaged 10 mg mL1 solution to
2.5 mg mL1 using sterile 0.9% preservative-free
saline (Baxter Healthcare Corp., IL, USA). All drugs
were prepared by one of the authors (RJB), who was
not involved in evaluating responses.
Two days before the study, each horse was stabled
individually in a stall to acclimatize. Access to grass
hay and water was provided at all times. On the day
before the study, each horse was instrumented with
an inflatable cuff at the base of the tail (width equal to
40% of the tail diameter) for noninvasive blood
pressure (NIBP) measurement (Cardell 9405; Minrad
87

International, Inc., NY, USA), and a chest strap heart
rate (HR) monitor (Polar; Polar Electro, Inc., NY,
USA), the accuracy of which was confirmed accord-
ing to facial artery pulse rate comparisons. Following
aseptic preparation of the skin and subcutaneous
infiltration of 1 mL of 2% lidocaine, a 14 gauge,
13.33 cm catheter (BD Angiocath; Becton Dickinson
Co., NJ, USA) was placed in each jugular vein to
facilitate drug administration and blood sampling,
respectively.
On the day of the study, two vials were delivered to
the treatment-blinded investigators (AC, SM). One
vial contained either propranolol or saline, and the
other vial contained either romifidine or saline. A
dose of 0.04 mL kg1 was removed from each vial,
combined in a 60 mL syringe with a 14 gauge needle,
and administered as a rapid IV bolus through the
catheter designated for drug administration. Blood
was collected from the catheter designated for blood
sampling before drug administration (time 0) and at
1, 2, 4, 8, 16, 30, 60, 120, 240, 480 and 960 mi-
nutes after treatment. Ten milliliters of blood were
discarded before each sample collection and the
catheter was flushed thereafter with heparinized
0.9% NaCl. Three milliliters of blood were collected in
syringes to which heparin was added for measure-
ment of packed cell volume (PCV), total protein (TP),
venous carbon dioxide and oxygen tensions (PvCO2
and PvO2, respectively), pH, base excess (BE), bicar-
bonate, total volume of carbon dioxide, hemoglobin,
saturation of hemoglobin with oxygen (SaO2),
lactate, glucose, sodium, chloride, potassium and
ionized calcium concentrations. Blood samples were
placed on ice immediately and analyzed within 1
hour of collection (ABL800 FLEX; Radiometer
America, Inc., OH, USA).
Samples of 10 mL of venous blood were collected,
placed in heparinized tubes and stored on ice until
measurement of drug plasma concentration (Cp).
Within 1 hour of collection, the blood was centri-
fuged at 6000 g for 5 minutes and the plasma
collected and immediately stored at 20  C until
analysis.
Drug response evaluation
Behavioral and physiologic responses were measured
and video-recorded at each sampling time point
(times 0e960 minutes) by two observers (AC, SM),
who were unaware of the drug treatments. Sedation,
a subjective evaluation of general activity and alert-
ness, was scored using a visual analogue scale (VAS)
88 © 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Romifidine and propranolol PK–PD in horses A Cenani et al.
consisting of a 100 mm line extending from 0 mm (no
sedation) to 100 mm (severe sedation). Tactile stim-
ulus responsiveness was assessed by slowly pressing a
10 mL syringe stopcock algometer (Johnson &
Watson 1997) against the neck of the horse, in
front of the shoulder about 10 cm below the crest.
The extent of plunger displacement in milliliters
required to cause the horse to move its neck, head,
ears or body was recorded. Visual responsiveness was
assessed using a modified technique from Ringer et al.
(2013). An open umbrella was moved toward the
horse's head at 0.3 m second1, starting 457 cm
away from it. The distance in centimeters at which
the horse appeared to notice the umbrella was
recorded. Ataxia was assessed using an ordinal scale
of 0e3: 0 (no sign of instability, stable); 1 (stable but
mild swaying); 2 (severe swaying or stumbling), and
3 (difficulty standing or falling down).
Vertical head height position from the ground to
the nose, respiratory rate ( fR) (assessed by counting
the number of thoracic excursions over a 15 second
period) and NIBP were recorded. HR measured with
the chest strap HR monitor was confirmed by
palpating facial artery pulses or by heart ausculta-
tion. Starting at 8 minutes after drug administration,
each abdominal quadrant was auscultated to score
gastrointestinal motility on a scale of 0e3: 0 (no
sound); 1 (1 borborygmus minute1); 2 (2 borbo-
rygmi minute1), and 3 (3 borborygmi minute1).
The scores from each quadrant were added to yield a
cumulative gastrointestinal motility score.
Determination of romifidine and propranolol
plasma concentrations
Romifidine (Toronto Research Chemicals, Inc., ON,
Canada) and propranolol (Cerilliant Corp., TX, USA)
analytical reference standards were combined into
one working solution for analysis. The working so-
lutions were prepared by dilution of the 1 mg mL1
stock solutions with methanol (Thermo Fisher Sci-
entific, Inc., NJ, USA) to concentrations of 0.01, 0.1,
1, 10 and 100 ng mL1. Plasma calibrators were
prepared by diluting the working standard solutions
with drug-free equine plasma to concentrations of
0.05e900 ng mL1. Fresh calibration curves and
negative control samples were prepared for each
quantitative assay. In addition, quality control sam-
ples (plasma fortified with analyte at three concen-
trations within the standard curve) were included
with each sample set as an additional check of
accuracy.
Elsevier Ltd. All rights reserved., 44, 86e97
Romifidine and propranolol PK–PD in horses A Cenani et al.
Prior to analysis, 500 mL of plasma was diluted with
200 mL of water (Honeywell, Burdick & Jackson
Laboratories, Inc., MI, USA) containing 25 ng mL1 of
d7-propranolol (Cerilliant Corp.) and d4-romifidine
(Toronto Research Chemicals, Inc.) internal stan-
dards, and 2 mL 0.6 M phosphate buffer (pH 6.5), and
vortexed briefly to mix. The samples were subjected to
solid phase extraction using Cerex PolyChrom Clin II
(35 mg in 3 mL) columns (Cera, Inc., CA, USA).
Samples were loaded onto and passed through the
columns using a Cerex System 48 Processor (Cera,
Inc.) with a positive pressure SPE manifold (SPEware
Corp., CA, USA). The columns were rinsed consecu-
tively with 3 mL water, 2 mL 1 M acetic acid,
and 3 mL methanol prior to elution with 3 mL
of methylene chloride:isopropranol:ammonium hy-
droxide (78:20:2 by v/v/v). Samples were dried under
nitrogen in a Zymark TurboVap (McKinley Scientific,
LLC, NJ, USA) at 45  C, reconstituted in 150 mL of 5%
acetonitrile (ACN) in water with 0.2% formic acid and
30 mL injected into the liquid chromatographyemass
spectrometry (LC-MS) system.
The concentration of romifidine and propranolol
was measured in plasma by LC tandem MS (LC-MS/
MS). Quantitative analysis of plasma was performed
using a TSQ Vantage triple quadrupole mass spec-
trometer (Thermo Scientific, Inc., CA, USA) coupled
with a turbulent flow chromatography system (TFC
TLX4; Thermo Scientific, Inc.) with an 1100 series LC
system (Agilent Technologies, Inc., CA, USA) and
operated in laminar flow mode.
Spray voltage was set at 3500 V, sheath gas and
auxiliary gas were 50 and 25 (arbitrary units),
respectively, vaporizer temperature was 350  C, and
the capillary temperature was 300  C. Product
masses and collision energies were optimized by
infusing the standards into the mass spectrometer.
Chromatography employed an ACE 3 C18 50.0  2.1
mm column (Mac-Mod Analytical, Inc., PA, USA)
and a linear gradient of ACN in water, both with
0.2% formic acid, at a flow rate of 0.25 mL minute1.
The initial ACN concentration was held at 65% for
0.33 minutes, ramped to 90% over 8.33 minutes,
and then re-equilibrated for 5.17 minutes at initial
conditions.
Detection and quantification were conducted using
selective reaction monitoring (SRM) of initial pre-
cursor ion for romifidine [mass to charge ratio (m/z)
258.0], propranolol (m/z 260.2), and the internal
standards d4-romifidine (m/z 262.0) and d7-
propranolol (m/z 267.2). Responses for the product
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
ions for romifidine (m/z 124.1, 159.1), propranolol
(m/z 155.1) and the internal standards d4-romifidine
(m/z 121.9, 126.1, 163.0, 164.1) and d7-
propranolol (m/z 161.2, 163.1, 188.1, 189.2) were
plotted and peaks at the proper retention time inte-
grated using Quanbrowser software (Thermo Scien-
tific, Inc.). Quanbrowser software was used to
generate calibration curves and to quantitate analy-
tes in all samples by linear regression analysis. A
weighting factor of 1/X was used for all calibration
curves.
Pharmacokinetic calculations
Compartmental analyses were used to determine
pharmacokinetic parameters in R, P and PR treat-
ments using commercially available software
(Phoenix WinNonlin Version 6.4; Pharsight Corp.,
NC, USA). The area under the curve (AUC) and area
under the moment curve were calculated using the
log-linear trapezoidal rule and extrapolation to in-
finity using the last measured Cp divided by the ter-
minal slope lz.
Statistical analysis
Data were visually assessed for normality. Once rela-
tive agreement between the VAS scores of the two
independent observers was confirmed using a Pearson
correlation coefficient of >0.7, values from observer
2 (SM) were transformed using PassingeBablok
regression to the scale used by observer 1 (AC) and
the two sets of data were then averaged.
Repeated-measures two-way analyses of variance
(ANOVAS) were used to detect differences in continuous
data between treatments and over time. Non-
parametric data were analyzed using the Friedman
test. HolmeSidak tests were used for parametric post
hoc pairwise comparisons, Wilcoxon rank-sum tests
were used for nonparametric post hoc comparisons.
All statistical analyses employed two-tailed tests
except for comparisons of HR between baseline and
between PR versus either P or R, comparisons of
scores for sedation, head ptosis and gastrointestinal
motility with baseline values, and clearance com-
parisons between PR versus either P or R. Because
these treatment effects were hypothesized a priori to
be altered in a single direction, one-sided tests were
used. A p-value of <0.05 was considered to indicate
differences of statistical significance. Data were
analyzed using STATA Version 12.0 (StataCorp LP, TX,
USA).
89

Results
Pharmacokinetic analyses
Pharmacokinetic parameters for IV administration of
R, P and PR are shown in Table 1 and Fig. 1. A two-
compartment model best fit the IV Cp versus time
curves of both R and P when drugs were adminis-
tered either together or separately. Drugs were
quantifiable in all horses at 960 minutes except in
two horses in R in which romifidine was detectable
only up to 480 minutes. Maximum Cp (Cmax) was
achieved at minutes 1 and 2 in P and R, respectively.
In PR, romifidine and propranolol Cmax was achieved
at minutes 1 and 2, respectively. Plasma concentra-
tions of both drugs decreased thereafter.
Pharmacodynamic changes
Baseline values collected before treatment adminis-
tration (minute 0) were within clinically acceptable
limits for adult horses. HR significantly decreased
compared with baseline in P, R and PR (Fig. 2).
Bradycardia, defined as a HR of <24 beats minute1,
was recorded in all horses in both R and PR, starting
as soon as 1 minute after drug administration.
Bradycardia was recorded in two horses in P, in both
of which an irregular rhythm with pulse deficits was
detected at baseline. There was a significantly
increased occurrence of HR measurements of <20
beats minute1 in PR compared with P (p ¼ 0.01),
with the lowest value of 12 beats minute1 detected
in PR in one horse at 1 minute after drug
administration.
Gastrointestinal motility significantly decreased
from baseline in R and PR at minutes 8e120 (p <
0.01). Gastrointestinal motility scores were not
changed in P. Ataxia was observed in R and PR, but
not in P (Fig. 3). Values for NIBP, fR and TP were
unchanged from baseline in all treatments. Overall,
treatments caused only clinically minor changes in
PCV, electrolyte and lactate concentrations and BE,
in comparison with baseline values, and values
remained within the institutional laboratory refer-
ence ranges for horses. PvO2 was significantly lower
than at baseline only in PR at minute 8 (p ¼ 0.02).
PvO2 was significantly lower in both R and PR than
in P at minutes 1e8 (p  0.01). The lowest PvO2
value detected was 27 mmHg (3.6 kPa) at minute 8
in both R and PR. Individual values for PvCO2 were
46e51 mmHg (6.1e6.8 kPa) in all treatments at all
time points. PvCO2 was higher in both R and PR than
in P at minute 16, whereas fR was unchanged in any
90 © 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Romifidine and propranolol PK–PD in horses A Cenani et al.
treatment. Glucose concentration significantly
increased compared with baseline in R at minute 16
(p ¼ 0.02), minutes 30e120 (p < 0.01) and minute
240 (p ¼ 0.01), and in PR at minutes 30e120 (p <
0.01) and minute 240 (p ¼ 0.01). Glucose concen-
tration was significantly higher in both R and PR
than in P at minutes 30e240 (p < 0.01).
Behavioral assessments over time showed similar
values for head height, tactile and visual assessments
in R and PR (Fig. 4). VAS scores over the first 16
minutes were significantly greater in PR than in R.
Sedation, ataxia and HR exhibited Cp-dependent
changes in which maximum effects were observed
at 100e200 ng mL-1, albeit under non-steady state
conditions (Fig. 5). At similar romifidine Cp, mean
behavioral changes were nonsignificantly greater in
PR than in R, particularly at the higher romifidine
concentrations (Fig. 5).
Discussion
For both romifidine and propranolol, the Cp time
course after an IV bolus was best described by a two-
compartment model, but with higher values for A, B
and AUC than previously reported (Aramaki et al.
2000; Wojtasiak-Wypart et al. 2012). By contrast
with these prior studies, both romifidine and pro-
pranolol were administered as an IV bolus and at
higher doses in the present study. This probably ex-
plains the higher peak plasma drug concentrations
reached in this study, as well as differences in some
pharmacokinetic parameters between the present
study and previous publications. Both romifidine and
propranolol are expected to dose-dependently
decrease CO (Muir & McGuirk 1985; Wojtasiak-
Wypart et al. 2012), thereby potentially reducing
liver blood flow and hepatic drug clearance (Branch
et al. 1973; Waterman 1984). Small errors in sam-
pling times during the first 1e2 minutes may also
have contributed to the higher distribution constant
and lower distribution half-life observed for pro-
pranolol compared with other results (Aramaki et al.
2000).
The propranolol dose in this study was selected
based on anecdotal use of a 1 mg kg1 oral dose by
local practitioners to treat anxiety or fear in horses.
The bioavailability of oral propranolol in horses is
about 4% and there is high variability among
animals (Aramaki et al. 2000). In the present study,
this dose was administered IV to ensure that phar-
macodynamic effects, either sedative or side effects,
did not go unobserved because of low Cp. In the study
Elsevier Ltd. All rights reserved., 44, 86e97
Elsevier Ltd. All rights reserved., 44, 86e97
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by

Romifidine and propranolol PK–PD in horses A Cenani et al.

Table 1 Pharmacokinetic parameters calculated in six horses for romifidine and propranolol after a single intravenous bolus of propranolol (1 mg kg1) (P), romifidine (0.1 mg kg1)
(R) or their combination (PR). Romifidine data were compared between R and PR; propranolol data were compared between P and PR

Romifidine Propranolol

Parameter R PR P PR

Vc(mL kg1) 463 ± 302; 352 (184e1036)
V2 (mL kg1) 1636 ± 449; 1662 (946e2338)
CI (mL kg1 minute1) 21 ± 4*; 20 (16e26)*
Cl2 (mL kg1 minute1) 23 ± 12; 21 (8e43)
ke (minute1) 0.05 ± 0.02; 0.06 (0.02e0.09)
k12 (minute1) 0.06 ± 0.04; 0.04 (0.03e0.13)
k21 (minute1) 0.01 ± 0.01; 0.01 (0.01e0.03)
a (minute1) 0.12 ± 0.05; 0.1 (0.1e0.2)
b (minute1) 0.006 ± 0.001; 0.005 (0.005e0.007)
A (ng mL1) 269 ± 150; 265 (83.5e532)
B (ng mL1) 16 ± 7; 14 (10e29)
AUC (ng minute mL1) 4940 ± 886*; 4969 (3901e6197)*
Cmax (ng mL1) 285 ± 150; 284 (96.5e543)
AUMC (ng minute2 mL1) 483,886 ± 101,549;
483,772 (332,066e628,828)
MRT (minute) 100 ± 22; 100 (70e132)
Vss (mL kg1) 2098 ± 720; 2029 (1130e3374)
t1/2a (minute) 7 ± 3; 6 (4e11)
t1/2b (minute) 121 ± 19; 127 (94e141)
370 ± 321; 275 (109e926)
1554 ± 244; 1519 (1290e1975)
17 ± 4; 17 (10e23)
22 ± 16; 19 (8e51)
0.07 ± 0.04; 0.07 (0.02e0.13)
0.14 ± 0.17; 0.10 (0.01e0.47)
0.01 ± 0.01; 0.01 (0.01e0.03)
0.22 ± 0.20; 0.19 (0.03e0.58)
0.005 ± 0.001; 0.005
(0.004e0.006)
483 ± 350; 464 (101e879)
18 ± 10; 19 (7e35)
6259 ± 2046; 5872
(4265e10,139)
502 ± 358; 485 (108e914)
831,182 ± 654,174;
632,404 (328,008e2,142,590)
121 ± 47; 111 (77e211)
1924 ± 278; 1917
(1540e2325)
7 ± 8; 4 (1e22)
143 ± 28; 120 (120e183)
1588 ± 411; 1665 (973e2140)
3798 ± 1462*; 3811
(1979e5741)*
25 ± 8*; 24 (16e35)*
91 ± 58; 68 (46e200)
0.016 ± 0.003*; 0.016
(0.011e0.020)*
0.07 ± 0.05; 0.04 (0.03e0.16)
0.03 ± 0.02; 0.02 (0.01e0.05)
0.1 ± 0.1; 0.1 (0.1e0.2)
0.004 ± 0.001; 0.004
(0.003e0.005)
520 ± 146; 502 (295e732)
152 ± 82; 149 (67e295)
42,630 ± 12,927*; 42,637
(28,510e62,575)*
672 ± 202; 602 (467e1028)
9,104,177 ± 3,061,970; 8,738,
110 (550,771e13,638,300)
213 ± 31; 205 (185e272)
5385 ± 1628*; 5244
(2952e7536)*
9 ± 4; 10 (3e13)
176 ± 26; 186 (137e200)
1699 ± 451; 1624 (1023e2399)
1371 ± 839; 1138 (495e2607)
14 ± 3; 15 (9e19)
44 ± 39; 43 (4e87)
0.009 ± 0.005; 0.009 (0.004e0.018)
0.030 ± 0.029; 0.025 (0.002e0.070)
0.04 ± 0.06; 0.02 (0.01e0.16)
0.08 ± 0.07; 0.06 (0.01e0.20)
0.004 ± 0.001; 0.004 (0.003e0.005)
396 ± 217; 391 (110e767)
233 ± 93; 215 (144e410)
74,003 ± 22,490; 67,276
(53,576e116,924)
628 ± 189; 616 (417e978)
17,182,733 ± 9,021,122;
13,321,400 (10,419,000e33,024,400)
222 ± 53; 205 (170e292)
3070 ± 534; 2925 (2416e3773)
25 ± 26; 17 (3e72)
180 ± 27; 183 (147e212)

Data are reported as both mean ± standard deviation and median (range). *Significantly different from PR (p < 0.05). Vc, volume of the central compartment; V2, volume of the
peripheral compartment; CI, systemic clearance; Cl2, CI of the peripheral compartment; ke, elimination constant from central compartment; k12, k21, transfer rate constant from the
central compartment to the peripheral compartment and vice versa; A and B are coefficients and a and b are first-order rate constants for the distribution (a) and elimination (b)
process; AUC, area under the plasma concentration curve; Cmax, peak concentration; AUMC, total area under the first moment curve; MRT, mean resident time; VSS, volume of
distribution at steady state; t1/2b, elimination half-life.
91
 Romifidine and propranolol PK–PD in horses A Cenani et al.

Figure 1 Drug plasma concentrations in one horse after intravenous administration of (a) propranolol (1 mg kg1; treat-
ment P) or propranolol and romifidine (0.1 mg kg1; treatment PR) and (b) romifidine (0.1 mg kg1; treatment R) or
romifidine and propranolol (PR).

by Aramaki et al. (2000), administration of pro-
pranolol (0.2 mg kg1) IV resulted in a peak Cp about
five times lower than that measured in our study.
Unfortunately Aramaki et al. (2000) reported only
propranolol pharmacokinetics and did not record
observations of sedation or side effects.
In the present study no major side effects or seda-
tion were observed in treatment P. Mild bradycardia
was observed in two of six horses in P. Pulse deficits,
which are likely to be attributable to second-degree
atrioventricular block, were detected in these two
horses at baseline prior to drug administration.

Figure 2 Heart rates in six horses before (baseline, minute 0) and after an intravenous bolus of propranolol (1 mg kg1;
treatment P), or romifidine (0.1 mg kg1; treatment R) or a combination of both at the same dose rates (treatment PR).
Values are mean ± standard deviation. * Significantly different from baseline (p  0.05). ySignificantly different from R and
PR (p  0.05).
92 © 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
Romifidine and propranolol PK–PD in horses A Cenani et al.

Figure 3 Ataxia scores in six horses before (baseline, minute 0) and after an intravenous bolus of romifidine (0.1 mg kg1;
treatment R) or romifidine (0.1 mg kg1) and propranolol (1 mg kg1) (treatment PR). Values are mean ± standard de-
viation. Score scale: 0, no sign of instability; 1, stable but swaying slightly; 2, clearly swaying; 3, falling down. *Significantly
different from baseline (p  0.05). ySignificantly different between treatments (p  0.05).

Profound bradycardia, with an increased occur-
rence of HR measurements of <20 beats minute1
and values as low as 12 beats minute1, was
observed in PR and was also associated with pro-
found ataxia and stumbling, which sometimes
required the observers to exit the stall. Potential risk
for fainting from low CO and cerebral blood flow
should be considered when bradycardia is as severe
as it was in some of the animals in PR. Severe and
progressive bradycardia may also result in cardiac
arrest and death. The highest reduction in HR from
baseline was observed at a romifidine Cp of about 50
ng mL1 in both the R and PR treatments. The degree
of HR reduction was more severe in PR than in R at
the same romifidine Cp. It is likely that propranolol
synergistically exacerbates romifidine-induced
bradycardia as propranolol alone at the same dose
had no significant HR effects.
Administration of a2-agonists results in brady-
cardia by a baroreceptor vagally mediated reflex
initiated by increased systemic vascular resistance
and decreased central sympathetic output (Rankin
2015). These sedatives also directly decrease HR by
inhibiting the cardiac hyperpolarization-activated
current that plays a role in the generation of pace-
maker potentials in sinoatrial node cells of the heart
(Knaus et al. 2007). Propranolol inhibits catechol-
amine binding at b-adrenergic receptors, slowing
spontaneous automaticity in the sinus node (Sage &
Mogg 2010). Both a2-agonists and b-blockers reduce
pacemaker activity and their simultaneous adminis-
tration can potentially increase the severity of
bradycardia.
The higher romifidine Cp values measured in PR
were most likely responsible for the more significant
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
ataxia seen in this treatment. Ataxia was most pro-
nounced at a romifidine Cp of about 70 ng mL1 in
both PR and R, but its severity was higher in PR than
in R at the same romifidine Cp. Propranolol alone
does not cause ataxia; therefore, exacerbation of
ataxia in PR could indicate a synergistic interaction
for this side effect, as previously noted for brady-
cardia. However, it is perhaps more likely that the
much slower HR observed in PR may have resulted in
a clinically significant reduction of CO and cerebral
blood flow that, in turn, may have produced addi-
tional weakness and ataxia.
In humans, propranolol produces anxiolysis by
blunting central sympathetic tone. Although many
anxiolytics are sedatives or potentiate actions of other
sedatives, propranolol did not cause sedation when
administered alone in the study horses. Although this
finding suggests a lack of tranquilization efficacy, the
present results may have been influenced by the
study population. Whereas propranolol is anxiolytic
in anxious people (Dyck & Chung 1991; Khadke et al.
2012), it does not produce behavior-altering benefits
in patients who do not suffer from anxiety disorders
(Jakobsson et al. 1995). It is possible that our results
may have been different if the studies had been con-
ducted in more stressed animals.
In treatment R, sedation was observed by 1 minute
after romifidine administration and persisted for 240
minutes. Onset was shorter and duration longer than
previously reported (Wojtasiak-Wypart et al. 2012).
Although these differences may be partially explained
by our use of a different sedation scoring method, the
higher romifidine Cp and the lower drug clearances in
the present study are most probably responsible for
the increased duration of sedation.
93
 Romifidine and propranolol PK–PD in horses A Cenani et al.

Figure 4 Behavioral responses in six horses before (baseline, minute 0) and after an intravenous bolus of propranolol (1 mg
kg1; treatment P), or romifidine (0.1 mg kg1; treatment R) or a combination of both at the same dose rates (treatment PR).
(a) Sedation visual analog scale (VAS). (b) Head height from the ground. (c) Response to a pressure algometer on the neck.
(d) Visual response to an approaching opened umbrella. In (c) and (d), minute 1 values are omitted owing to missing data.
Values are mean ± standard deviation. *Significantly different from baseline (p < 0.05). ySignificantly different from R and PR
(p < 0.05). zSignificantly different from PR (p < 0.05).

94 © 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
Romifidine and propranolol PK–PD in horses A Cenani et al.

Figure 5 Romifidine plasma concentrations (Cp) in six horses before (baseline, minute 0) and after an intravenous bolus of
romifidine (0.1 mg kg1; treatment R) or romifidine (0.1 mg kg1) and propranolol (1 mg kg1) (treatment PR) compared
with: (a) sedation visual analog scale (VAS); (b) ataxia score, and (c) heart rate (HR) expressed as the percentage of baseline
values, where baseline is 100% and represents HR before drug administration at a romifidine Cp of 0 ng mL1. HR values at
romifidine Cp values of 0.1e0.2 ng mL1 were 95% and 104% of baseline in R and PR, respectively; values omitted for figure
clarity.

Co-administration of propranolol and romifidine decreased and the AUC increased for both romifidine
doubled the duration of sedation compared with and propranolol when administered together versus
romifidine administered alone at the same dose. separately. This potentially may be explained by a
Sedation magnitude was also greater in PR versus R for decrease in hepatic blood flow and drug metabolism
16 minutes after drug administration. Both of these (Branch et al. 1973; Ochs et al. 1980; Waterman
effects mirrored the higher romifidine Cp achieved in 1984), and a decrease in VC (central volume of distri-
the PR treatment compared with R. Drug clearance bution) caused by low CO induced by the combination
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by 95
Elsevier Ltd. All rights reserved., 44, 86e97

of propranolol and romifidine. Furthermore, propran-
olol can directly reduce hepatic metabolism of
drugs independently of hepatic blood flow (Greenblatt
et al. 1978; Bowdle et al. 1987). It is unknown
whether this can occur with romifidine in horses.
Arguably, there is little benefit to be gained by
coadministering propranolol and romifidine to pro-
duce the deeper sedation caused by higher romifidine
Cp. A more logical approach would be to simply
administer more romifidine to achieve these desired
concentrations. When evaluated as a function of
romifidine Cp, sedation VAS scores peaked and
reached a plateau at about 40 ng mL1. In a previ-
ously published study, horses were considered suffi-
ciently sedated at romifidine Cp of about 30 ng mL1
(Ringer et al. 2012). In the present study, a moderate
level of sedation appropriate for standing chemical
restraint (sedation VAS scores of 6e7) in both the R
and PR treatments corresponded to a romifidine Cp of
around 40 ng mL1. Our pharmacokinetic data can
be used to calculate dose rates for a loading dose (LD)
and constant rate infusion (CRI) to reach and main-
tain a romifidine Cp close to 40 ng mL1. The LD can
be calculated by the following equation: LD ¼ Cp ∙ VD,
where VD is the volume of distribution. If the mean of
VC and VSS (volume of distribution at steady state)
(equal to 1280 mL kg1) is used to estimate VD, the
romifidine LD equals 51 mg kg1. Using VC or VSS
alone to estimate VD may result in Cp values that are
either far below or far above the target drug concen-
tration, and fail to reach the pharmacologic end point
required or increase adverse effects or toxicity,
respectively (Wada et al. 1998). A CRI to maintain
this plasma concentration is calculated as: CRI ¼ Cp ∙
Cl, where Cl is the romifidine clearance (21 mL kg1
minute1). This yields a romifidine CRI equal to 0.84
mg kg1 minute1. It is unknown whether a further
increase in romifidine Cp would be beneficial given the
side effects observed at higher levels of Cp in this study.
Interestingly, the VAS scores for sedation were
higher in PR than R at the same romifidine Cp. This
suggests that despite the increase in romifidine Cp,
propranolol may have potentiated the sedative effect
of romifidine through a non-pharmacokinetic mech-
anism. As the same sedative effect was not seen when
propranolol was administered alone, the severe
bradycardia in PR cannot be excluded as a possible
cause by potentially decreasing HR and CO and
consequently cerebral blood flow, thereby reducing
the degree of consciousness.
Gastrointestinal side effects commonly reported with
propranolol administration in humans are diarrhea
96 © 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Romifidine and propranolol PK–PD in horses A Cenani et al.
and cramps (Horvath et al. 1979; Abbott et al. 2010).
Although romifidine predictably decreased gastroin-
testinal motility, the other side effects were not
observed in the study horses. Propranolol may cause
hypoglycemia as a result of blockade of hepatic glyco-
genolysis and by attenuation of an adrenergically
induced increase in blood glucose concentration (Perry
2007; Horev et al. 2015). Hyperglycemia in response
to a2-agonists in horses is well documented (Kullmann
et al. 2014) and was observed in both R and PR. The
higher blood glucose concentrations in PR than in R
may reflect the higher romifidine Cp in PR.
Conclusions
Propranolol (1 mg kg1) IV failed to produce detect-
able sedation or a decreased response to tactile and
visual stimuli in the horses in this study. Compared
with romifidine alone, the combination of propranolol
and romifidine doubled the duration of sedation and
increased the intensity of sedation for the first 16
minutes, but the combination treatment was also
associated with more severe bradycardia and ataxia.
The severity of these side effects does not support the
co-administration of these drugs for the sedation of
horses at the doses studied.
Authors' contributions
AC: study design, the acquisition and analysis of data,
and the writing of the manuscript. RJB: contributed
to the conception of the project, the study design, and
the acquisition and analysis of data. SM: the acqui-
sition of data. HKK: contributed to drug analysis and
pharmacokinetic modeling. JEM: contributed to the
conception of the project. All authors contributed to
the critical revision of the paper and approved the
final manuscript for publication.
References
Abbott J, Parulekar M, Shahidullah H et al. (2010) Diar-
rhea associated with propranolol treatment for hem-
angioma of infancy (HOI). Pediatr Dermatol 27, 558.
Aramaki S, Mori M, Nakata M et al. (2000) Pharmaco-
kinetics of propranolol and its metabolites in horses
after intravenous or oral administration. Biol Pharm
Bull 23, 1333e1340.
Bowdle TA, Freund PR, Slattery JT (1987) Propranolol re-
duces bupivacaine clearance. Anesthesiology 66, 36e38.
Branch RA, Shand DG, Wilkinson GR, Nies AS (1973)
The reduction of lidocaine clearance by dl-propranolol:
an example of hemodynamic drug interaction.
J Pharmacol Exper Ther 184, 515e519.
Elsevier Ltd. All rights reserved., 44, 86e97
Romifidine and propranolol PK–PD in horses A Cenani et al.
Doze V, Chen BX, Li Z, Maze M (1989) Pharmacologic
characterization of the receptor mediating the hypnotic
action of dexmedetomidine. Acta Vet Scand 85(Suppl.),
61e64.
Dyck JB, Chung F (1991) A comparison of propranolol
and diazepam for preoperative anxiolysis. Can J
Anaesth 38, 704e709.
England GC, Clarke KW, Goossens L (1992)
A comparison of the sedative effects of three a2-adre-
noceptor agonists (romifidine, detomidine and xyla-
zine) in the horse. J Vet Pharmacol Ther 15, 194e201.
Etkin A, Wager TD (2007) Functional neuroimaging of
anxiety: a meta-analysis of emotional processing in
PTSD, social anxiety disorder, and specific phobia. Am J
Psychiatry 164, 1476e1488.
Freeman SL, England GC (2001) Effect of romifidine on
gastrointestinal motility, assessed by transrectal ultra-
sonography. Equine Vet J 33, 570e576.
Gasthuys F, Parmentier D, Goossens L, De Moor A (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, 489e502.
Greenblatt DJ, Franke K, Huffman DH (1978) Impair-
ment of antipyrine clearance in humans by propran-
olol. Circulation 5 (7), 1161e1164.
Horev A, Haim A, Zvulunov A (2015) Propranolol induced
hypoglycemia. Pediatr Endocrinol Rev 12, 308e310.
Horvath F Jr, Marbury TC, Uhlemann ER et al. (1979)
Severe diarrhea secondary to propranolol. South Med J
72, 367e368.
Hurlemann R, Walter H, Rehme AK et al. (2010) Human
amygdala reactivity is diminished by the b-noradrenergic
antagonist propranolol. Psychol Med 40, 1839e1848.
Jakobsson J, Rane K, Ryberg G (1995) Oral premed-
ication one hour before minor gynaecological surgery
e does it have any effect? A comparison between
ketobemidone, lorazepam, propranolol and placebo.
Acta Anaesthesiol Scand 39, 359e363.
Johnson TW, Watson PJ (1997) An inexpensive, self-
assembly pressure algometer. Anaesthesia 52,
1070e1072.
Khadke VV, Khadke SV, Khare A (2012) Oral propranolol
e efficacy and comparison of two doses for peri-operative
anxiolysis. J Indian Med Assoc 110, 457e460.
Knaus A, Zong X, Beetz N et al. (2007) Direct inhibition
of cardiac hyperpolarization-activated cyclic
nucleotide-gated pacemaker channels by clonidine.
Circulation 115, 872e880.
Kullmann A, Sanz M, Fosgate GT et al. (2014) Effects of
xylazine, romifidine, or detomidine on hematology,
biochemistry, and splenic thickness in healthy horses.
Can Vet J 55, 334e340.
Moens Y, Lanz F, Doherr MG, Schatzmann U (2003)
A comparison of the antinociceptive effects of xylazine,
detomidine and romifidine on experimental pain in
horses. Vet Anaesth Analg 30, 183e190.
© 2016 Association of Veterinary Anaesthetists and American College of Veterinary Anesthesia and Analgesia. Published by
Elsevier Ltd. All rights reserved., 44, 86e97
Muir WW, McGuirk SM (1985) Pharmacology and phar-
macokinetics of drugs used to treat cardiac disease in
horses. Vet Clin North Am Equine Pract 1, 335e352.
Ochs HR, Carstens G, Greenblatt DJ (1980) Reduction in
lidocaine clearance during continuous infusion and by
coadministration of propranolol. N Engl J Med 303,
373e377.
Perry PJ (2007) Antianxiety agents (anxiolytics). In:
Psychotropic Drug Handbook (8th edn). Perry PJ,
Alexander B, Liskow BI, DeVane CL (eds). Lippincott
Williams & Wilkins, USA. pp. 296e328.
Rankin DC (2015) Sedatives and tranquilizers. In: Lumb
& Jones' Veterinary Anesthesia and Analgesia (5th
edn). Grimm KA, Lamont LA, Tranquilli WJ et al.
(eds.). Wiley-Blackwell, UK. pp. 196e206.
Ringer SK, Portier KG, Fourel I, Bettschart-Wolfensberger R
(2012) Development of a romifidine constant rate infu-
sion with or without butorphanol for standing sedation of
horses. Vet Anaesth Analg 39, 12e20.
Ringer SK, Portier K, Torgerson PR et al. (2013) The
effects of a loading dose followed by constant rate
infusion of xylazine compared with romifidine on
sedation, ataxia and response to stimuli in horses. Vet
Anaesth Analg 40, 157e165.
Sage A, Mogg T (2010) Pharmacology of drugs used to treat
cardiac disease. In: Cardiology of the Horse (2nd edn).
Marr CM, Bowen M (eds). Saunders, USA. pp. 75e87.
Strange BA, Dolan RJ (2004) b-Adrenergic modulation of
emotional memory-evoked human amygdala and hip-
pocampal responses. Proc Natl Acad Sci USA 101,
11454e11458.
Wada DR, Drover DR, Lemmens HJ (1998) Determina-
tion of the distribution volume that can be used to
calculate the intravenous loading dose. Clin Pharma-
cokinet 35, 1e7.
Wang C, Knowles MG, Chakrabarti MK, Whitwam JG
(1994) Clonidine has comparable effects on sponta-
neous sympathetic activity and afferent Ad and C-fiber-
mediated somatosympathetic reflexes in dogs. Anes-
thesiology 81, 710e717.
Waterman AE (1984) The pharmacokinetics of ketamine
administered intravenously in calves and the modi-
fying effect of premedication with xylazine hydrochlo-
ride. J Vet Pharmacol Ther 7, 125e130.
Wojtasiak-Wypart M, Soma LR, Rudy JA et al. (2012)
Pharmacokinetic profile and pharmacodynamic effects
of romifidine hydrochloride in the horse. J Vet Phar-
macol Ther 35, 478e488.
Woodhouse KJ, Brosnan RJ, Nguyen KQ et al. (2013)
Effects of postanesthetic sedation with romifidine or
xylazine on quality of recovery from isoflurane anes-
thesia in horses. J Am Vet Med Assoc 242, 533e539.
Received 3 December 2015; accepted 8 March 2016.
Available online 23 February 2017
97