Content uploaded by Butch Kukanich
Author content
All content in this area was uploaded by Butch Kukanich on Jan 29, 2019
Content may be subject to copyright.
JAVMA, Vol 243, No. 1, July 1, 2013 Scientific Reports 105
EQUINE
The use of analgesics in horses has been evolving over
the past few decades. As veterinarians and horse
owners have become more aware of the importance of
pain management, there has been a push toward use of
multimodal analgesia. Morphine is a µ-opioid agonist
that was formerly used routinely in equine practice as
a preanesthetic and analgesic agent. However, the use
of morphine in equine practice has waned because of
the potential for adverse effects. Gastrointestinal stasis1
and CNS excitation2 are among the described adverse
effects of morphine administered at high doses. New-
er classes of drugs, such as α-2 receptor agonists and
agonist-antagonist opioids, are associated with fewer
adverse effects and have begun to take the place of mor-
phine in routine practice. As the importance of analge-
sia has become more readily accepted, administration
of NSAIDs has become common in equine practice.
Pharmacokinetics of intramuscularly
administered morphine in horses
Elizabeth P. Devine, DVM, MS, DACVS; Butch KuKanich, DVM, PhD, DACVCP; Warren L. Beard, DVM, MS, DACVS
Objective—To determine the pharmacokinetics of morphine after IM administration in a
clinical population of horses.
Design—Prospective clinical study.
Animals—77 horses.
Procedures—Morphine sulfate (0.1 mg/kg [0.045 mg/lb], IM) was administered to horses,
and blood samples were obtained at predetermined time points. Plasma morphine concen-
trations were measured via liquid chromatography and mass spectrometry. In preliminary
investigations, samples were obtained from 2 healthy horses at 12 time points (up to 12
hours after drug administration) and analyzed via 2-stage pharmacokinetic analysis. In the
clinical phase, blood samples were obtained from 75 hospitalized horses at various times
(total, 2 to 3 samples/horse) up to 9 hours after drug administration, and data were analyzed
via a naïve pooled pharmacokinetic model.
Results—In the clinical phase, the apparent terminal half-life (t1/2) of morphine was approxi-
mately 1.5 hours, volume of distribution per bioavailability was approximately 4.5 L/kg, and
clearance per bioavailability was approximately 35 mL/kg/min. Peak plasma concentration
in naïve pooled analysis was 21.6 ng/mL and occurred approximately 4 minutes after ad-
ministration. Morphine concentrations were below the limit of quantification ≤ 7 hours after
administration in 74 horses. Adverse effects attributed to morphine administration were
uncommon and considered mild.
Conclusions and Clinical Relevance—The short t1/2 of morphine in horses suggested fre-
quent administration may be needed to maintain targeted plasma concentrations. Varia-
tions in plasma concentrations suggested optimal dosages may differ among horses. The
drug was well tolerated at the described dose, but patients receiving morphine should be
monitored carefully. (J Am Vet Med Assoc 2013;243:105–112)
However, the use of these medications is not without
consequence; complications of NSAID administration
include right dorsal colitis, gastric ulceration, and re-
nal failure. Despite these effects, NSAIDs are widely
used because of their anti-inflammatory and analgesic
properties. For some conditions, including laminitis,
infected synovial structures, and fractures, the degree
of analgesia obtained with NSAIDs alone is inadequate,
and multimodal analgesia, such as the combined use of
opioids and NSAIDS, can be beneficial.
Most pharmacokinetic studies typically involve a
small number of healthy horses (eg, 4 to 8) and frequent
blood sample collection to determine the absorption,
distribution, metabolism, and excretion of the drug in
each member of a homogenous group. The mean, me-
dian, and range of pharmacokinetic parameters are re-
ported. These studies, termed standard 2-stage studies,
are useful for estimation of mean parameters in a well-
defined, homogenous population. However, a small,
From the Departments of Clinical Sciences (Devine, Beard) and Anat-
omy and Physiology (KuKanich), College of Veterinary Medicine,
Kansas State University, Manhattan, KS 66502.
This manuscript represents a portion of a thesis submitted by Dr.
Devine to the Kansas State University Department of Biomedical
Sciences as partial fulfillment of the requirements of a Master of
Science degree.
Supported in part by the Department of Clinical Sciences Research
Committee Funding and the Analytical Pharmacology Laboratory
at Kansas State University.
Address correspondence to Dr. Devine (devine@vet.k-state.edu).
ABBREVIATIONS
Cl Clearance
Cl/F Clearance per bioavailability
LOQ Limit of quantification
t1/2 Apparent terminal half-life
Vd Volume of distribution
Vd/F Volume of distribution per bioavailability
106 Scientific Reports JAVMA, Vol 243, No. 1, July 1, 2013
EQUINE
homogenous group of healthy animals may not be rep-
resentative of the clinical population of horses that may
benefit from morphine as an analgesic.
Several other methods for assessing the pharmaco-
kinetics of a drug are available, including population
pharmacokinetics and naïve pooled pharmacokinet-
ics. In contrast to standard 2-stage pharmacokinetic
studies, population and naïve pooled pharmacokinetic
studies may involve large numbers of animals, often
including the targeted clinical population, and do not
require frequent sample collection from individual pa-
tients. Population pharmacokinetics includes a larger,
more diverse group, compared with standard 2-stage
studies, and allows biological variability to be evalu-
ated. Factors affecting variability in a veterinary popu-
lation may include age, breed, body weight, concurrent
medications, and health status.3,4 Population pharma-
cokinetics allows identification of patient traits that
may alter the pharmacokinetics of a drug; however, one
of the disadvantages of this type of study is that fitting
a model to the data can be very difficult with some data
sets.
Naïve pooled pharmacokinetic studies are often
used to minimize sample collection from individual pa-
tients and to allow for evaluation of a group of patients.
They include large numbers of individuals and can be
performed in the targeted clinical population. In con-
trast to population pharmacokinetics, a single model is
fit to pooled data from all individual patients. There-
fore, factors such as age, breed, and body weight cannot
be assessed for effects on pharmacokinetic parameters.
However, naïve pooled pharmacokinetic studies typi-
cally do not require the specialized software needed for
population pharmacokinetics, and a single model is fit
to all of the data, making the modeling less complex.
Although morphine has been used for many years
in horses, pharmacokinetic parameters for the drug are
not well described in this species. The dose and fre-
quency of IM administration of morphine have been
empirical, selected on the basis of the practitioner’s
experience and on the degree of discomfort evidenced
by clinical signs in the horse. Because adverse effects
can result from excessive morphine administration, it
is imperative that a proper dose and administration fre-
quency be clarified to minimize these effects. The first
step in identifying an appropriate dose is to determine
pharmacokinetic parameters of the drug following ad-
ministration via the desired route. The purpose of the
study reported here was to determine the pharmacoki-
netics of morphine after IM administration in a clini-
cal population of horses. This population was chosen
to allow an assessment of the variability in the plasma
concentrations of morphine in patients receiving the
drug as premedication for anesthesia or as treatment
for signs of pain instead of in healthy research animals.
Materials and Methods
Animals—Two investigator-owned healthy adult
Quarter Horses, a 9-year-old gelding and a 17-year-old
mare, were used for initial evaluation of plasma mor-
phine concentrations and to determine the LOQ of the
drug prior to the clinical phase of the study. These
horses had no abnormalities detected via physical ex-
amination and were known to have been healthy for
4 and 6 years. For the clinical portion of the study, all
horses admitted to the Kansas State University Veteri-
nary Medical Teaching Hospital from December 2010
to June 2011 were eligible for enrollment. Horses were
included if morphine was administered as a periop-
erative analgesic or as treatment for signs of pain. All
horses were treated according to their clinical signs
and diagnosis as deemed necessary by the supervising
clinician. Pregnant mares and horses known to have
impaired pulmonary, hepatic, or renal function or en-
docrine disorders were not included in the study. The
preliminary investigation and the clinical study were
approved by the Kansas State University Institutional
Animal Care and Use Committee and performed in
accordance with that committee’s guidelines. Prior to
enrollment of client-owned horses in the study, own-
ers were informed of the nature of the study and pro-
vided their consent.
Preliminary investigation—Both horses were indi-
vidually housed and had access to food and water dur-
ing the preliminary investigation. The skin over the left
jugular vein was aseptically prepared, and an IV cath-
eter was placed to allow for serial blood sample collec-
tion. The 2 horses were weighed, and morphine sulfatea
(0.1 mg/kg [0.045 mg/lb], IM) was administered in the
left side of the neck at time 0. Whole blood samples
(7 mL) were collected into heparinized tubes at 5, 10,
15, 30, and 45 minutes and 1, 2, 4, 6, 8, 10, and 12
hours after morphine administration. The IV catheter
was flushed with heparinized saline (0.9% NaCl) so-
lution after each blood sample was collected and was
removed after the last sample was obtained. Paired
(duplicate) sets of blood samples were obtained from 1
horse (randomly selected via coin toss).
Physical examinations were performed at the time
of morphine administration and every 2 hours thereaf-
ter for 12 hours to monitor for any potential adverse ef-
fects. In each examination, heart rate, respiratory rate,
and rectal temperature were measured; the abdomen
was auscultated for presence of borborygmus; defeca-
tions were noted; the injection site was monitored for
heat and swelling; demeanor was assessed; and signs of
sedation (mild [lack of interest in the surroundings],
moderate [ataxic when walked], or severe [recumbent
or unresponsive]) or excitation (mild, [restlessness in
the stall], moderate [constant walking or pacing], or se-
vere [attempts to climb in the stall or other potentially
dangerous behavior]) were recorded. Horses were ad-
ditionally monitored for signs of colic or other adverse
effects.
Blood samples were centrifuged for 5 minutes at
5,000 X g, and the plasma was frozen ≤ 2 hours after
sample collection. The duplicate blood samples ob-
tained from 1 horse were stored on ice until the end of
the day and centrifuged together at that time. There-
fore, samples collected at the beginning of the day were
on ice for just over 12 hours before the serum was
separated, and the last sample obtained was centrifuged
within 1 hour after collection. This was done to deter-
mine whether sample handling would affect results.
All plasma samples were stored in a freezer at –70°C
prior to analysis. Morphine concentrations were deter-
JAVMA, Vol 243, No. 1, July 1, 2013 Scientific Reports 107
EQUINE
mined by means of liquid chromatography and mass
spectrometry.
Clinical phase—Horses were weighed, and mor-
phinea (0.1 mg/kg, IM) was administered in the neck
at time 0. The side of the neck used for morphine ad-
ministration varied among horses and was chosen to
avoid sites where other medications had been previous-
ly administered. Two or 3 blood samples (7 mL each)
were collected from each horse at different times after
morphine administration until approximately 9 hours
after injection. Samples were obtained via jugular ve-
nipuncture or from an IV catheter, when present. The
time of sample collection was recorded to the minute,
and sample collection times were evenly distributed
throughout the 9-hour collection period. After collec-
tion, samples were refrigerated at 4°C until centrifuga-
tion. Blood samples were centrifuged as previously de-
scribed ≤ 8 hours after collection, and the plasma was
separated and stored in a freezer at –70°C until analy-
sis. Physical examinations were performed at the time
of morphine administration, every 2 hours thereafter
(until the time point established in the preliminary in-
vestigation at which morphine was no longer detect-
able in samples), and at the time of each blood sample
collection. Examinations included monitoring the same
variables as described for the preliminary investigation;
evaluations were not performed when a horse was in
surgery or recovering from surgery, or if the patient was
discharged from the hospital before a scheduled exami-
nation. Concurrent medications received, including
anesthetic agents and IV fluids, were recorded for each
patient along with breed, age, body weight, sex, and di-
agnosis. Feed was withheld for 6 to 8 hours prior to
anesthesia for horses that were anesthetized for surgery.
Water was not withheld.
Analysis of plasma morphine concentrations—
Liquid chromatography and mass spectrometry were
used for plasma sample analysis. Samples were thawed
on a heat block at 40°C and centrifuged at 5,000 X g
for 5 minutes. Hydromorphone D3b (internal standard;
0.1 mL; 100 ng/mL) and borate buffer (1 mL; 0.1M; pH,
9.2) were added to 1 mL of plasma, and the sample was
mixed via vortexing for 5 seconds.
The C18 solid-phase extraction cartridgesc were at-
tached to a solid-phase extraction manifold and then
conditioned with 1 mL of methanol followed by 1 mL
of deionized water. The plasma, hydromorphone, and
buffer mixtures were run through the conditioned car-
tridges attached to the solid-phase extraction manifold,
and then the cartridges were rinsed with 1 mL of deion-
ized water each. One milliliter of methanol was added
to each cartridge to elute the sample. The methanol
was evaporated to dryness at 40°C under an air stream
for 30 minutes. Each sample was then resuspended in
0.2 mL of 50% methanol in deionized water and mixed
thoroughly. The sample was centrifuged at 15,000 X g
for 5 minutes, and the supernatant was removed and
placed in injection vials.
Plasma morphine standards were processed at least
twice daily in an identical manner to that used for sam-
ples to ensure consistency of the assay. The interday ac-
curacy of the assay determined on replicates of 5 at con-
centrations of 2.5, 10.0, and 50.0 ng/mL were 102%,
98%, and 99% of the actual concentration, respectively.
The interday coefficients of variation determined on
replicates of 5 at concentrations of 2.5, 10.0, and 50.0
ng/mL were 9%, 9%, and 7%, respectively. The analytic
LOQ was 2.5 ng/mL, defined as the lowest concentra-
tion of the standard curve with measured concentra-
tions within 15% of the actual concentration.
Pharmacokinetic analysis—Only plasma samples
with morphine concentrations greater than the LOQ
were included in the pharmacokinetic analysis. A
1-compartment open model with first-order input and
output with no lag was used for analysis of samples
from the preliminary investigation and for naïve pooled
pharmacokinetic analysis in the clinical phase. Popula-
tion pharmacokinetic modeling was assessed via a soft-
ware program with a nonlinear mixed-effects model.d
One- and 2-compartment models with first-order input
and output with and without lag were assessed. The
primary pharmacokinetic parameters estimated were
the Vd, absorption rate constant, and elimination rate
constant. Secondary pharmacokinetic parameters in-
cluded the apparent absorption half-life, t1/2, area un-
der the curve, Cl, maximum plasma concentration, and
time to the maximum plasma concentration.
Statistical analysis—A Friedman repeated-mea-
sures ANOVA on ranks test was used to analyze heart
rate and respiratory rate data. Comparison between
the duplicate samples was performed by calculating
the difference between samples centrifuged ≤ 2 hours
after collection and those batched for later centrifuga-
tion. Then, a linear regression model was used to de-
termine a relationship between the samples centrifuged
≤ 2 hours after collection and those centrifuged up to
12 hours after collection. Statistical softwaree was used
for data analysis. Values of P < 0.05 were accepted as
significant for all statistical tests.
Results
Preliminary investigation—Measured variables
evaluated in the physical examinations remained with-
in respective reference ranges throughout the 12-hour
evaluation period after morphine administration in
both horses. There was no change in behavior score
observed, and no evidence of a local injection site reac-
tion was detected in either horse. Signs of gastrointes-
tinal motility were considered normal and no signs of
colic were detected. One horse defecated between the
2- and 4-hour time points and again between the 4- and
6-hour time points. The other horse defecated between
the 4- and 6-hour time points and again between the
8- and 10-hour time points. No adverse effects were ob-
served during the study period.
Peak plasma morphine concentrations in the 2
horses were estimated at 13 and 24 minutes, with
concentrations of 33.5 and 28.2 ng/mL, respectively.
Concentrations were above the LOQ (2.5 ng/mL) for
8 hours in one horse and 6 hours in the other. For du-
plicate samples obtained from 1 horse to evaluate the
effects of sample handling on drug concentrations, val-
ues were compared between the paired samples and the
percentage difference between samples collected at the
108 Scientific Reports JAVMA, Vol 243, No. 1, July 1, 2013
EQUINE
same time points was determined. On the basis of lin-
ear regression analysis with a weighting factor of 1/y2,
where y is the plasma drug concentration, the R2 value
of 0.991 (P < 0.001) indicated a strong linear relation-
ship between the samples centrifuged for plasma sepa-
ration at different times (1 to 6 hours after collection;
correlation was not evaluated at later time points be-
cause of the low morphine concentrations), suggesting
drug concentrations in samples remained stable for up
to 6 hours after collection.
Clinical phase—Seventy-five horses were in-
cluded in the clinical phase of the study. One horse, a
10-month-old Quarter Horse colt, was included twice
because it received morphine on 2 separate visits to the
clinic but was only included once in signalment-related
data. Quarter Horses (52/75 horses) were overrepre-
sented. Other breeds or breed types included warm-
bloods (n = 4), American Paint Horse (4), draft breeds
(2), Tennessee Walking Horse (2), Thoroughbred (2),
Arabian (2), mixed (2), and Morgan, Saddlebred, Mus-
tang, Rocky Mountain Spotted Horse, and Pony of the
Americas (1 each). Eighteen mares, 22 geldings, and
35 stallions were included. The mean age was 6.2 years
(range, 7 months to 31 years). Body weights ranged
from 204 to 896 kg (449 to 1,971 lb), with a mean of
415 kg (913 lb).
Seventeen of 75 horses were treated for orthopedic
abnormalities. Six of these horses did not undergo sur-
gery; these included horses with laminitis, foot abscess,
severe osteoarthritis, laceration that extended into a
joint space, and fractured sacral and caudal vertebrae.
The remaining 11 horses with orthopedic abnormali-
ties underwent surgical procedures, including tarso-
crural, middle carpal joint, and metatarsophalangeal
joint arthroscopy; right hind proximal interphalangeal
joint (pastern) arthrodesis; repair of an ulnar fracture;
removal of the apical fragment of a fractured sesamoid
bone; desmotomy of the accessory ligaments of the
deep and superficial digital flexor tendons; and seques-
trum removal from a right second metacarpal bone.
Thirty-four horses (including 4 that were cryptorchid)
underwent castration; the horse that received morphine
twice during the study had surgery postponed on the
first visit and underwent castration on the second visit.
Fifteen horses underwent soft tissue surgery other than
castration, which included surgery of the upper respi-
ratory tract, treatment of postoperative colic, repair of
a rectovaginal tear, palmar digital neurectomy, neuro-
ma removal, abdominal hernia repair, and esophageal
diverticulum repair. Three horses underwent surgical
treatment of oral or facial conditions, including incisor
removal, repair of a mandibular fracture, and frontona-
sal sinus flap surgery for ventral conchal sinusitis. Four
horses underwent ocular surgery, including removal of
a third eyelid, tumor (periocular squamous cell carci-
noma) resection, and enucleation following rupture of
a corneal ulcer. The final category included 2 horses
with soft tissue lacerations. In total, 59 horses under-
went procedures requiring general anesthesia.
Of 76 patients that received morphine during this
part of the study, 74 also received phenylbutazone or
flunixin meglumine. Because plasma morphine con-
centrations were not reliably detectable after the 8-hour
time point in the preliminary investigation, monitor-
ing of the described physical examination variables
was discontinued after 8 hours in the clinical phase of
the study. In addition, 25 horses were discharged from
the hospital before the 8-hour time point, reducing the
number of data points included in subsequent analyses.
In clinical patients, the median respiratory rate
was 20 breaths/min at time 0 and ranged from 18 to 24
breaths/min after morphine administration (Table 1).
There was no difference in respiratory rate among time
points (P = 0.742). Median heart rate was 44 beats/min
at time 0 and ranged from 40 to 44 beats/min afterward.
Median heart rate at 8 hours (40 beats/min) was signifi-
cantly (P = 0.028) lower than that at 4 hours (44 beats/
min), but no other significant differences were detected
for this variable.
Fecal output was recorded every 2 hours for each
horse for the duration of hospitalization; however,
discharge of 25 horses prior to the 8-hour time point
likely resulted in some defecations not having been re-
corded in this interval. Of 17 horses that did not un-
dergo general anesthesia, 2 defecated before the 2-hour
time point, 3 defecated between the 2- and 4-hour time
points, and 4 defecated between the 4- and 6-hour time
points. The remaining horses either did not defecate
by 8 hours (n = 2) or were discharged before 8 hours
(6). Among the 59 horses that underwent general an-
esthesia, the first defecation was within 2 hours after
morphine administration for 8 horses, between 2 and
4 hours for 4 horses, and between 4 and 6 hours for 17
horses; 9 horses defecated for the first time between the
6- and 8-hour time points. Ten horses evaluated for 8
hours did not defecate during that time, and 11 horses
were not observed to defecate but left the clinic prior to
the 8-hour time point.
Mild excitation was observed in 1 horse 4 hours af-
ter morphine administration. This was a colt admitted
for castration that had not been halter trained and had
not previously been confined in a stall. Sedation was
observed in 31 horses; however, in most (29) of these,
sedation was observed immediately after return to the
stall following anesthesia and surgery or after admin-
istration of an α-2 receptor agonist (xylazine or deto-
midine) or phenothiazine tranquilizer (acepromazine).
Two horses with signs of sedation had not received oth-
er medications. Signs of mild sedation were observed
Respiratory rate Heart rate
Time No. of horses (breaths/min) (beats/min)
0 76 20 (12–66) 44 (28–80)
2 53 24 (12–100) 44 (28–80)
4 59 18 (12–60) 44 (32–72)
6 56 18 (8–72) 43 (28–64)
8 51 20 (12–44) 40 (28–64)
Data were collected from 75 horses, with 1 horse included twice
on separate visits. Decrease in the number of horses over time re-
flects discharge of patients at various times after drug administra-
tion. The time of morphine injection was considered time 0.
Table 1—Median (range) heart and respiratory rates for 76 horses
that received morphine (0.1 mg/kg, IM, once) as a perioperative
analgesic or as treatment for signs of pain in the clinical phase of
a study to evaluate pharmacokinetics of the drug.
JAVMA, Vol 243, No. 1, July 1, 2013 Scientific Reports 109
EQUINE
in one of these horses at the 6- and 8-hour time points
and in the other horse at the 4- and 6-hour time points.
Two horses had injection site reactions; neither re-
quired treatment. One horse had a 2.5-cm, focal, cir-
cumscribed swelling at the injection site that was de-
tected at the 4-hour time point. Palpation of the region
did not elicit signs of pain. This condition persisted to
the end of the 8-hour observation period but was not
evident on the following day. The other horse had a
1.5-cm swelling detected at the 4-hour time point that
was also not associated with signs of pain on palpa-
tion. At the 8-hour time point, the size of the swell-
ing had decreased by approximately half, and other as-
pects were unchanged; the condition had completely
resolved by the next morning.
Blood samples (2 or 3/horse) were collected at
various time points throughout the day following mor-
phine administration. Twenty-six samples were collect-
ed during the first hour, 26 were collected from 1 to < 2
hours, and 22 were collected from 2 to < 3 hours. The
plasma morphine concentration was above the LOQ in
all 74 samples obtained < 3 hours after drug admin-
istration and in 49 of 51 samples obtained between 3
and 5 hours after injection. Thirteen of 48 samples had
detectable concentrations of morphine between 5 and 7
hours after injection, and only 1 of 44 samples met this
criterion after 7 hours.
Pharmacokinetic analysis—An attempt was made
to fit the clinical data to a population pharmacokinetic
model, but this could not be accomplished satisfactorily.
Therefore, a naïve pooled population analysis method
was used to determine the pharmacokinetic parameters
of morphine after IM injection in these horses (Table
2). In the preliminary experiments and in the clinical
portion of the study, morphine administered IM had a
rapid absorption phase followed by a slower elimina-
tion phase (Figure 1). Healthy horses in the prelimi-
nary investigation had higher peak plasma concentra-
tions of morphine (28.2 and 33.5 ng/mL), compared
with the value for horses of the clinical population
evaluated via naïve pooled pharmacokinetic analysis
(21.6 ng/mL). Although individual results for healthy
horses in the preliminary investigation indicated higher
maximum drug concentrations than that determined
for the clinical population, results for those 2 horses
fit within the range of values for clinical patients (11.7
to 48.7 ng/mL). The absorption rate constant for mor-
phine in clinical patients was 79.76 hours–1, compared
with a mean value of 12.95 hours–1 for the 2 horses in
the preliminary investigation. The apparent half-life of
the absorption phase had a mean value of 3.7 minutes
in horses in the preliminary investigation and was only
30 seconds in the clinical population. Maximal con-
centrations of morphine in horses in the preliminary
investigation were detected at a mean of 19 minutes,
whereas peak concentrations were detected in clinical
patients at approximately 4 minutes after administra-
tion. The t1/2 was 1.48 hours in clinical patients, with a
mean of 1.76 hours in healthy horses of the preliminary
investigation.
Because IV administration of morphine was not in-
cluded in our study, bioavailability could not be calcu-
lated. Therefore, Vd and Cl were calculated as Vd/F and
Cl/F rather than absolute values. The Vd/F in horses of
the preliminary investigation had a mean value of 2.9 L/
kg and was 4.49 L/kg for clinical patients evaluated via
naïve pooled pharmacokinetic analysis. The Cl/F had a
mean value of 19.05 mL/kg/min for horses in the pre-
liminary investigation and was 34.9 mL/kg/min for the
clinical population.
Discussion
Morphine has been used as an analgesic in horses for
years but has dose-related adverse effects. Interestingly, a
Preliminary
investigation
Clinical
Parameter Mean Range phase
Ka (h–1) 12.95 8.21–17.68 79.76
Apparent absorption half-life (h) 0.0618 0.0392–0.0844 0.00869
Kel (h–1) 0.397 0.361–0.432 0.467
t1/2 (h) 1.76 1.6–1.96 1.48
Vd/F (L/kg) 2.9 2.72–3.07 4.49
AUC (h•ng/mL) 87.7 85–90.3 47.74
Cl/F (mL/min/kg) 19.05 18.5–19.6 34.9
Cmax (ng/mL) 30.85 28.2–33.5 21.6
Tmax (h) 0.31 0.22–0.4 0.065
In preliminary investigations, samples were collected at 5, 10,
15, 30, and 45 minutes and 1, 2, 4, 6, 8, 10, and 12 hours after mor-
phine administration and data were analyzed via standard 2-stage
pharmacokinetic analysis. In the clinical phase, samples were ob-
tained from hospitalized horses at various times (total, 2 to 3 sam-
ples/horse) up to 9 hours after drug administration and data were
analyzed via a naïve pooled pharmacokinetic model.
Cmax = Maximum plasma concentration. Ka = Absorption rate
constant. Kel = Elimination rate constant. Tmax = Time of maximum
plasma concentration.
See Table 1 for remainder of key.
Table 2—Pharmacokinetic parameters of morphine (0.1 mg/kg, IM, once)
following administration to 2 healthy horses during preliminary investiga-
tions and to 76 equine patients during the clinical phase of the study.
Figure 1—Plasma drug concentrations (semilogarithmic scale) of
morphine following administration in horses (0.1 mg/kg, IM). The
time of morphine injection was considered time 0. Mean values
obtained from 2 healthy horses during preliminary investigations
(triangles) and individual values obtained from 76 clinical patients
(data from 75 horses, with 1 horse included twice on separate vis-
its) that received morphine as a perioperative analgesic or as treat-
ment for signs of pain (circles) are shown. The solid line indicates the
predicted plasma kinetic profile of morphine determined via naïve
pooled pharmacokinetic analysis of samples from clinical patients.
The dashed line indicates the predicted profile determined via stan-
dard 2-stage pharmacokinetic analysis in preliminary investigations.
110 Scientific Reports JAVMA, Vol 243, No. 1, July 1, 2013
EQUINE
number of studies1,5–8 have been performed to evaluate
various doses of morphine administered via IV and IM
routes and their clinical effects in horses, but the phar-
macokinetic parameters for morphine administered IM
in horses have not been reported. In a previous study,9
morphine (0.1 mg/kg) administered IV in horses had a
t1/2 of 1.6 hours, and mean plasma concentrations of the
drug were < 10 ng/mL within 4 hours and < 5 ng/mL
within 6 hours after administration. A 3-compartment
model was used to describe pharmacokinetics of mor-
phine in that study.9 In another study in horses,10 phar-
macokinetic analysis of morphine (0.25 mg/kg [0.11 mg/
lb]) administered IV during general anesthesia with iso-
flurane revealed a t1/2 of 0.7 hours, Vd of 1.2 L/kg, and
Cl of 40 mL/min/kg. In the same study,10 the t1/2 of mor-
phine following IV administration at a dose of 2.0 mg/
kg (0.91 mg/lb) was 1 hour. Those authors10 postulated
that the t1/2 of morphine following IV administration in
horses is dose dependent. Administration of a medica-
tion into a compartment other than the central venous
compartment can alter its parameters in plasma because
of differences in absorption, and subsequently, the t1/2 of
extravascular morphine may be substantially different
following administration via different routes. Additional-
ly, poor absorption from extravascular routes may result
in low and possibly subtherapeutic drug concentrations.
Therefore, it is important to assess the pharmacokinetic
parameters of morphine following IM administration in
horses to evaluate the possibility of whether it could be a
clinically useful route of administration.
In the present study, the pharmacokinetic param-
eters of morphine administered IM as a perioperative
analgesic or as treatment for signs of pain in a clinical
population of horses included a t1/2 of approximately
1.5 hours, Vd/F of 4.49 L/kg, and Cl/F of 34.9 mL/min/
kg. The values of Vd/F and Cl/F may be useful for other
calculations involving morphine administered IM in
horses, but these do not imply 100% bioavailability of
the drug.
A 1-compartment open model with first-order in-
put and output with no lag was used for pharmacoki-
netic analysis of samples from 2 healthy horses in our
preliminary investigation and for naïve pooled pharma-
cokinetic analysis of samples from a clinical population
(75 horses, with 1 patient that received morphine on 2
visits for a total of 76). The model was chosen on the
basis of visual inspection of the data and the model that
best fit the data from the preliminary investigation. Al-
though plasma concentrations below the LOQ (2.5 ng/
mL) could not be assessed, morphine was likely to be
present in the plasma of horses after the last time point
used in the present study. We can speculate that if drug
concentrations below the LOQ in our study could be
quantified with a more sensitive assay, they would re-
sult in uniform residuals on the terminal portion of the
curve, suggesting a 1-compartment model is the most
appropriate model for the data set used in our naïve
pooled pharmacokinetic analysis. However, it is pos-
sible that a 2-compartment model could best describe
this data set if the assay used was sensitive enough to
quantify concentrations lower than the LOQ.
Peak plasma morphine concentrations following
IM administration in both healthy horses in the pre-
liminary investigation were within the range of con-
centrations detected in the clinical population, despite
numeric differences in some of the pharmacokinetic
parameters. A substantial degree of variability was de-
tected in plasma morphine concentrations after IM ad-
ministration in the clinical population of horses. It is
important to consider this variability, recognizing that
1 dose may not have the same effects in all patients,
and to perform routine clinical monitoring of equine
patients that receive morphine IM to maximize drug ef-
ficacy and minimize adverse effects.
Some apparent differences in pharmacokinetic pa-
rameters of morphine were detected between healthy
horses in the preliminary investigation and the larger
population of horses that received morphine in the
clinical phase of the study. Sample collection during the
preliminary investigation was most intensive during
the drug absorption and elimination phases, whereas
samples collected for naïve pooled analysis were pri-
marily obtained during the elimination phase. This was
done intentionally to completely capture the elimina-
tion phase because this is more clinically important for
determination of dosing intervals than is the absorption
phase. However, these data may incompletely represent
the absorption phase, thus biasing absorption rate and
time of maximal drug concentration in the naïve pooled
analysis, which was a limitation of our study. The small
number of samples during the absorption phase may
also have contributed to the inability to fit a population
pharmacokinetic model to the data.
Reported IV doses for morphine in horses range
from 0.02 to 2.4 mg/kg (0.009 to 1.09 mg/lb) in experi-
mental and clinical settings.8,11 Some doses have been
associated with adverse effects in horses, including de-
creased gastrointestinal motility and CNS excitation. At
a dose of 1.0 mg/kg (0.45 mg/lb) IV, morphine slowed
the passage of feces in the gastrointestinal tract of hors-
es.1 Other investigators reported a significant decrease
in fecal moisture content and a significant increase in
gastrointestinal transit time in horses after repeated
doses of morphine (0.5 mg/kg [0.23 mg/lb], IV, q 12 h)
in a crossover-design study.7 Horses had a 22-hour de-
lay in passing barium-impregnated spheres after mor-
phine treatment, compared with the values following
saline solution (control) treatment, and signs of CNS
stimulation were detected for 2 to 4 hours after mor-
phine administration (0.5 mg/kg, IV) in 3 of 5 horses.7
In experiments conducted to determine the morphine
dose used in that study,7 3 treatments of morphine (1.0
mg/kg, IV) caused severe signs of colic in 1 of 2 horses.
In other investigations, morphine administered IM at
0.66 mg/kg (0.30 mg/lb) caused restlessness in 7 of 8
ponies from 1 to approximately 4 hours after admin-
istration,5 and horses became ataxic and appeared un-
aware of their surroundings following a dose of 2.4 mg/
kg, IV.11 In the present study, a dose of 0.1 mg/kg, IM,
appeared to have little effect on heart rate and degree of
excitation, and no signs of colic were observed in any
horses. In a retrospective study2 to evaluate adverse ef-
fects in horses that received 0.1 to 0.17 mg of morphine/
kg (0.045 to 0.077 mg/lb), IV, the authors concluded
that morphine administration did not increase the risk
of intraoperative or postoperative complications. This
JAVMA, Vol 243, No. 1, July 1, 2013 Scientific Reports 111
EQUINE
dose range has been used clinically as an analgesic at
other institutions7 and is used for IM administration of
morphine at the hospital where the present study was
performed.
Morphine was only consistently detectable with
our assay for 3 hours after IM administration in clinical
patients. After that time, the plasma morphine concen-
tration in some horses was < 2.5 ng/mL. Because sever-
al samples lacked detectable morphine concentrations
during the later time points in the study, it is likely that
frequent administration is necessary to provide consis-
tent analgesia in some clinical patients when the drug
is administered via this route.
Determination of the efficacy of morphine at the
dose and route used was beyond the scope of the pres-
ent study. Pharmacodynamic studies evaluating asso-
ciations between pharmacokinetic data and analgesia
have not been reported in horses, and this is an area
for potential further research. Pharmacodynamic stud-
ies of morphine have been performed in humans, and
although this information cannot be extrapolated di-
rectly to horses, it does provide some general guidelines
to help equine practitioners to make educated deci-
sions about dosage recommendations. In a prospective
study12 of human cancer patients with chronic pain,
substantial individual variability was detected in the se-
rum concentrations of morphine needed for analgesia.
The range of effective plasma concentrations was 30 to
120 nmol/L (8.58 to 34 ng/mL) for the 25th to 75th
percentiles. When patient-controlled analgesia was as-
sessed in the immediate postoperative period, there was
also a large degree of individual variability, with a mean
± SD minimum effective plasma concentration of 16 ±
9 ng/mL.13 If 16 ng of morphine/mL is used as an ap-
proximate value for the mean minimum analgesic con-
centration in horses, only 18 of 26 (69%) samples col-
lected between 0 to 1 hours and 9 of 26 (35%) samples
collected between 1 and 2 hours had morphine concen-
trations that exceeded this value.
Although the model used in the present study can-
not predict the safety of other doses of morphine in
horses, it can be used to predict the plasma concentra-
tions of morphine after the IM administration at other
doses. According to our model, if this population of
horses received morphine IM at a dose of 0.2 mg/kg
(0.09 mg/lb) and the plasma concentrations are pro-
portional to the dose, all 26 samples collected between
0 and 1 hours after injection and 20 of 26 samples
obtained between 1 and 2 hours after injection would
contain > 16 ng of morphine/mL. Two to 3 hours after
drug administration, 8 of 22 samples collected would
have morphine concentrations above this value. Al-
though dose-related changes in circulating morphine
concentrations can be extrapolated with this model,
other studies1,7 have shown that increasing the dose of
morphine may increase the number of adverse effects.
Therefore, studies should be conducted assessing the
pharmacokinetic parameters and adverse effects of
morphine at higher doses than that used in the pres-
ent study. Marked variability was detected in plasma
morphine concentrations in clinical patients in our
study. This model does suggest that higher doses
would be needed to achieve targeted plasma morphine
concentrations in a larger percentage of horses and
that higher doses would likely extend the duration for
which concentrations of morphine exceed 16 ng/mL.
Clinical judgment and careful patient monitoring is
required in determining the appropriate dose to maxi-
mize analgesia and minimize adverse effects.
Morphine use for patient-controlled analgesia was
measured in a human study13 with morphine as the
only analgesic administered. Although administration
of 1 type of analgesic can be effective, the use of > 1
class of analgesics can also be beneficial. Multimodal
analgesia combines analgesics with different mecha-
nisms of action in an attempt to enhance analgesia
and minimize the adverse effects of a single medica-
tion.14,15 A systematic review16 of 60 human studies
concluded that the amount of morphine required for
analgesia was reduced by the concurrent administra-
tion of an NSAID. There was a substantial decrease in
postoperative nausea and vomiting, which are among
the adverse effects of morphine in humans, in patients
that received an NSAID as well as morphine.16 There
is clear evidence that multimodal analgesia is effec-
tive in human medicine, and at least 1 clinical report14
suggests the same is true in horses. In the present
study, almost every horse in the clinical population
that received morphine (74/76) also received either
phenylbutazone or flunixin meglumine. The effects
of concurrent administration of an NSAID on mor-
phine analgesic effects have not been evaluated in a
population of horses. It is likely that combined drug
treatments in horses will decrease the plasma concen-
tration of morphine needed for analgesia, but further
study is needed to determine effective concentrations
of morphine when used alone or in combination with
other medications.
One limitation of our study was that the observer
assessing sedation and excitation was not blinded,
and this could have biased these evaluations. Anoth-
er limitation is that it is difficult to draw any mean-
ingful conclusions about fecal output because most
(59/76) horses underwent general anesthesia and had
feed withheld for 6 to 8 hours before surgery. Even
without opioid administration, general anesthesia
has been reported to inhibit gastrointestinal motility
in horses.17 Most (47/76) horses in the present study
defecated within 8 hours after morphine administra-
tion and, of the 29 horses that were not observed to
defecate during this time, 17 were discharged before
the end of the 8-hour observation period. Therefore,
it seems reasonable to conclude that if morphine
administered at this dose inhibited gastrointestinal
motility, the effect was transient. However, it is im-
portant to note that associations between higher or
multiple doses of morphine and gastrointestinal mo-
tility were not evaluated.
Most horses (59/76) underwent general anesthesia
and surgery on the same day that morphine was admin-
istered. To our knowledge, a comparison of the phar-
macokinetic parameters of morphine with and without
concurrent general anesthesia has not been reported.
Similarly, the effects of health status, breed, and sex on
these values have not been reported in horses. There-
fore, numerous factors could have influenced the re-
112 Scientific Reports JAVMA, Vol 243, No. 1, July 1, 2013
EQUINE
sults of pharmacokinetic analysis of morphine in this
study. These data were obtained in clinical patients and
therefore should be more representative of the phar-
macokinetics of the targeted population than data ob-
tained in a study of healthy horses.
Injection site reactions were detected in only 2
horses in the present study. Both horses had a localized
swelling at the injection site that did not seem pain-
ful on palpation and resolved without treatment. These
swellings may have represented hematomas, seromas,
or allergic reactions. However, because the reactions
were self-limiting and transient, no other tests were
performed to determine the nature of the swelling. We
are unaware of any other reports describing injection
site reactions after IV or IM administration of morphine
in horses. Although median heart rate in clinical pa-
tients was significantly lower at 8 hours (40 beats/min)
than at 4 hours (44 beats/min) after drug administra-
tion, both values were within the reference range for
horses (28 to 44 beats/min). Therefore, the difference
did not appear to have any clinical relevance.
Time from sample collection to centrifugation and
freezing varied during the clinical phase of our study,
but data collected during the preliminary investigation
indicated that time from collection to centrifugation of
the sample under similar storage conditions resulted
in minimal differences in the measured plasma con-
centration of morphine. This was important because
it can be challenging to obtain multiple blood samples
from a large number of clinical patients and perform
plasma separation immediately after collection. Our
preliminary data showed that separation of plasma up
to 6 hours after collection had minimal effects on the
amount of morphine that was detected in the plasma.
Although further testing is needed, results of the
present study indicate that plasma concentrations of
morphine equivalent to concentrations considered ef-
ficacious in humans are transient following administra-
tion of a single 0.1 mg/kg dose IM in horses. Because
of the potential for adverse effects at higher or more
frequent doses, care should be taken to monitor equine
patients receiving morphine for evidence of efficacy or
adverse effects and to use the lowest dose that has ther-
apeutic benefit for the patient.
a. Morphine sulfate, Baxter Healthcare Corp, Deerfield, Ill.
b. Cerilliant Inc, Round Rock, Tex.
c. Bond Elut, Varian Inc, Palo Alto, Calif.
d. WinNonMix, version 2.0.1, Pharsight Corp, Cary, NC.
e. SigmaStat, version 3.1, Systat Software Inc, Chicago, Ill.
References
1. Alexander F. The effect of some anti-diarrhoeal drugs on intes-
tinal transit and faecal excretion of water and electrolytes in the
horse. Equine Vet J 1978;10:229–234.
2. Mircica E, Clutton RE, Kyles KW, et al. Problems associated
with perioperative morphine in horses: a retrospective case
analysis. Vet Anaesth Analg 2003;30:147–155.
3. Cox SR, Liao S, Payne-Johnson M, et al. Population pharma-
cokinetics of mavacoxib in osteoarthritic dogs. J Vet Pharmacol
Ther 2011;34:1–11.
4. Peyrou M, Doucet MY, Vrins A, et al. Population pharmacoki-
netics of marbofloxacin in horses: preliminary analysis. J Vet
Pharmacol Ther 2004;27:283–288.
5. Kalpravidh M, Lumb WV, Wright M, et al. Effects of butorpha-
nol, flunixin, levorphanol, morphine, and xylazine in ponies.
Am J Vet Res 1984;45:217–223.
6. Phaneuf LP, Grivel ML, Ruckebusch Y. Electromyoenterography
during normal gastro-intestinal activity, painful or non-painful
colic and morphine analgesia, in the horse. Can J Comp Med
1972;36:138–144.
7. Boscan P, Van Hoogmoed LM, Farver TB, et al. Evaluation of the
effects of the opioid agonist morphine on gastrointestinal tract
function in horses. Am J Vet Res 2006;67:992–997.
8. Clark L, Clutton RE, Blissitt KJ, et al. The effects of morphine
on the recovery of horses from halothane anaesthesia. Vet An-
aesth Analg 2008;35:22–29.
9. Combie JD, Nugent TE, Tobin T. Pharmacokinetics and protein
binding of morphine in horses. Am J Vet Res 1983;44:870–874.
10. Steffey EP, Eisele JH, Baggot JD. Interactions of morphine and
isoflurane in horses. Am J Vet Res 2003;64:166–175.
11. Combie J, Dougherty J, Nugent E, et al. Pharmacology of nar-
cotic analgesics in the horse IV. Dose and time response rela-
tionships for behavioral-responses to morphine, meperidine,
pentazocine, anileridine, methadone, and hydromorphone. J
Equine Med Surg 1979;3:377–385.
12. Klepstad P, Dale O, Kaasa S, et al. Influences on serum con-
centrations of morphine, M6G and M3G during routine clini-
cal drug monitoring: a prospective survey in 300 adult cancer
patients. Acta Anaesthesiol Scand 2003;47:725–731.
13. Dahlström B, Tamsen A, Paalzow L, et al. Patient-controlled an-
algesic therapy, part IV: pharmacokinetics and analgesic plas-
ma concentrations of morphine. Clin Pharmacokinet 1982;7:
266–279.
14. Dutton DW, Lashnits KJ, Wegner K. Managing severe hoof pain
in a horse using multimodal analgesia and a modified composite
pain score. Equine Vet Educ 2009;21:37–43.
15. Wideman GL, Keffer M, Morris E, et al. Analgesic efficacy of a
combination of hydrocodone with ibuprofen in postoperative
pain. Clin Pharmacol Ther 1999;65:66–76.
16. Maund E, McDaid C, Rice S, et al. Paracetamol and selective
and non-selective non-steroidal anti-inflammatory drugs for the
reduction in morphine-related side-effects after major surgery:
a systematic review. Br J Anaesth 2011;106:292–297.
17. Singh S, McDonell WN, Young SS, et al. Cardiopulmonary and
gastrointestinal motility effects of xylazine/ketamine-induced
anesthesia in horses previously treated with glycopyrrolate. Am
J Vet Res 1996;57:1762–1770.
