Pharmacokinetics of butorphanol in horses after intramuscular injection


Debra C. Sellon, Department of Veterinary Clinical Sciences, Washington State University, Pullman, WA 99164, USA. E-mail:


A two-way cross-over study of the pharmacokinetics of butorphanol after intravenous and intramuscular administration at 0.08 mg/kg in six adult horses was performed. Heparinized venous blood samples were obtained prior to drug administration and at 10, 20, 30, 45, 60, 120, 180, 240, and 360 min after IV injection. Samples were obtained at the same time points and at 6 h and 12 h after IM injection. Physical examination parameters were recorded at each time point. Plasma butorphanol concentrations were determined by high performance liquid chromatography. No significant differences in any physical parameters were observed after butorphanol administration except for an increase in respiratory rate at 60 and 180 min after IV administration. Absorption of butorphanol after IM administration was very rapid (half life of absorption of 6 min) but systemic availability after IM injection was low (37%). Terminal half-life after IV administration was much longer than half-life after IM administration (0.57 h and 7.7 h, respectively). This difference was attributed to detection of a deep compartment after IV administration that was not detectable after IM administration. To maintain targeted plasma butorphanol concentrations above 10 ng/mL, administration of 0.08 mg/kg IM every 3 h may be necessary.


The most commonly used analgesic medications in horses are nonsteroidal anti-inflammatory drugs such as phenylbutazone and flunixin meglumine. These drugs are nonspecific inhibitors of cyclo-oxygenase (COX) enzymes and exert their effects by blocking prostaglandin synthesis at sites of inflammation. Use of these drugs may be associated with a variety of adverse effects including gastric and colonic ulceration, renal papillary necrosis and inhibition of gastrointestinal epithelial restitution following injury (Moses and Bertone, 2002; Tomlinson and Blikslager, 2003). Attempts to decrease exposure by adjusting the dose of these drugs in order to minimize the potential for adverse effects often occur at the expense of maintaining adequate levels of analgesia.

Opiate analgesic agents are often used in other species as part of a plan for multimodal analgesia to diminish postoperative pain (Kehlet, 1991; Kehlet, 1989). We have previously shown that administration of a continuous rate infusion (CRI) of butorphanol, a narcotic agonist-antagonist, results in steady-state plasma concentration of drug within the predicted therapeutic range (Sellon et al., 2001). Administration of CRI butorphanol to horses in combination with intravenous flunixin meglumine after exploratory celiotomy results in more normal behavior, decreased cortisol concentrations, less weight loss and more rapid discharge from the hospital (Sellon et al., 2004). Although CRI is efficacious, it is not always practical for treatment of horses with potentially painful disorders, especially if those horses are not already receiving intravenous fluid therapy. Therefore, we sought to determine the pharmacokinetics of butorphanol after intramuscular injection to facilitate design of efficacious multimodal analgesic protocols for horses.

Materials and methods

Six healthy adult horses, four castrated male and two female, were used for this study. Body weight ranged from 409 to 577 kg. Horses were housed in box stalls and allowed free access to grass hay throughout treatment and observation periods. All experimental protocols were approved by the Washington State University Institutional Care and Use Committee. A two-way crossover trial was performed to determine the pharmacokinetics of intramuscular butorphanol in horses. Horses were treated with intravenous (n = 3) or intramuscular (n = 3) butorphanol at 0.08 mg/kg. After a 7 week wash-out period treatments were reversed. Prior to administration of butorphanol, a catheter was placed in the left jugular vein of each horse. A catheter was also placed in the right jugular vein for drug administration in horses receiving intravenous butorphanol. Intramuscular injections were administered in the muscle on the left side of the neck. Depending on body weight, each horse received a single intramuscular injection of 3.2 to 4.6 mL total volume.

Blood samples were collected from the left jugular vein catheter by aspiration with a syringe prior to drug administration (time 0) and at 10, 20, 30, 45, 60, 120, 180, 240, and 360 min after IV injection. Samples were similarly obtained at the same time points and at 6 h and 12 h after IM injection. Blood samples were immediately injected into a heparinized glass tube and stored on ice until centrifuged for collection of plasma. Plasma was stored at −80 °C until butorphanol concentrations were measured. Physical examination data (general behavior and demeanor, heart rate, respiratory rate, gastrointestinal sounds, and passage of manure) were recorded for each horse at the same time points that blood samples were obtained. General behavior and demeanor was described subjectively as normal or abnormal. Gastrointestinal auscultation score was calculated as previously described (Sellon et al., 2001). Physical examination data (heart rate, respiratory rate, gastrointestinal auscultation score) from each horse were compared to baseline (time 0) data from that horse using a one way analysis of variance for repeated measures.

Plasma butorphanol concentrations were determined by high performance liquid chromatography (HPLC) as previously described (Sellon et al., 2001). The only change in the assay was an update in equipment to the Agilent 1000 system (Agilent Technologies Inc, Palo Alto, CA 94306, United States) that allowed a more stable baseline and improved lower limit of quantification and detection.

Plasma concentrations of butorphanol after a single IV or IM injection were analyzed using a computer program (WinNonlin version 5.0, Pharsight, Mountian View, California). Plasma concentration vs. time curves were first plotted to aid in selection of the most appropriate model for analysis. For the IM injection, the model selected was a one-compartment model with first-order input. For the IV data, the model selected was a two-compartment model with elimination from the central compartment. These models were tested against other models using a Goodness of Fit criteria to select the best statistical model using the Akaike’s Information Criteria (AIC) described by Yamaoke et al. (1978). Pharmacokinetic calculations were performed using the approaches and formulae presented in other sources (Gibaldi and Perrier, 1982; Gabrielsson and Weiner, 2001). The general formula for the IM model was:


where C(T) is the plasma concentration at time = T, D is dose, K01 is the absorption rate constant, K10 is the terminal rate constant, e is the base of the natural logarithm, and V is the apparent volume of distribution. Secondary parameters calculated for the IM dose were half-life of absorption, terminal half-life, area-under-the-curve (AUC), and peak concentration (CMAX).

The extent of absorption (F) was calculated from a comparison of the AUC values from IM and IV (F = AUCIM/AUCIV).

For the IV dose, a two-compartment model was selected that corresponded to the general formula:


where A and B are the intercepts for the distribution and elimination rate constants, respectively. Secondary parameters calculated included the absorption and elimination rate half-lives, apparent volumes of distribution, systemic clearance, mean residence time (MRT), microdistribution rate constants (K10, K12, K21), and AUC.


There was no significant difference in heart rate over time after administration of IV or IM butorphanol. At baseline, mean heart rates were 45.0 ± 3.0 beats/min in horses receiving IV butorphanol and 42.0 ± 2.2 beats/min in horses receiving IM butorphanol. Although there was no difference in respiratory rate over time after IM butorphanol (P = 0.712), respiratory rate was significantly increased from baseline (15.5 ± 2.7 breaths/min) at 60 and 180 min (23.7 ± 9.5 and 23.3 ± 6.4 breaths/min, respectively) after IV drug administration. There was no significant difference in gastrointestinal auscultation score over time after administration of butorphanol by IV or IM route. At baseline, mean gastrointestinal auscultation score was 3.3 ± 1.6 in horses receiving IV butorphanol and 2.5 ± 1.8 in horses receiving IM butorphanol. Behavior of all horses was subjectively considered normal at each time point during the experimental period.

Mean values for pharmacokinetic parameters after IV and IM butorphanol are shown in Table 1. Absorption of butorphanol after IM administration was very rapid (half life of absorption was only 6 min) but systemic availability of the IM injection was low (37%).

Table 1.   Pharmacokinetic of butorphanol after intravenous (IV) or intramuscular (IM) administration of a single dose of 0.08 mg/kg to 6 healthy adult horses
  1. For the IM dose: VD/F, apparent volume of distribution corrected for absorption; K01, absorption rate constant; K10, terminal rate constant; AUC, area-under-the-curve; CL/F, systemic clearance corrected for absorption; Tmax, time of peak concentration; Cmax, peak concentration; %F, percent systemic absorption.

  2. For the IV dose: A and B, distribution and elimination y-axis intercepts; α and β, distribution and elimination rate constants; K10, K12, and K21, intercompartmental transfer rate constants; Vc, apparent volume of distribution of the central compartment; Vss, apparent volume of distribution at steady-state; CL, systemic clearance.

 Absorption half-lifeh0.100.04
 Terminal half-lifeh0.5710.141
 α half lifeh0.300.06
 β half lifeh7.775.05

For two horses, the intravenous data was difficult to fit; therefore, for these two horses the limits of the terminal slope of the curve were controlled in the final analysis in order to meet reasonable expectations.


There were several differences between the results of this study and our previously published study of pharmacokinetics of butorphanol after IV bolus and continuous rate infusion (Sellon et al., 2001). The assay used for this current study is more sensitive than previously used assays. Because improved equipment was available, limits of detection were as low as 1 or 2 ng/mL compared to approximately 5 to 6 ng/mL in the previous report. This increased assay sensitivity may have resulted in an ability to detect a longer elimination half-life after IV bolus injection in the current study (7.8 h as compared to 44 min, respectively) because we can detect levels for a longer time. Evidence for this is that the terminal half-life reported in the earlier study is closer to what is reported here for the IM dose. We also reported in this study a difference in terminal half-life between IM and IV administration (0.57 h and 7.7 h, respectively). The IV injection produced a long terminal slope (Fig. 1) because the sensitive assay could detect plasma concentrations from the IV injection longer. Because the absorption was low after IM injection, these late terminal points could not be detected using our assay. In other words, because concentrations after IV injection were initially so much higher than the IM injection, we were able to detect a deep compartment that was not detected in the IM study. A more sensitive assay of the samples obtained after IM drug administration may have been able to detect this deep compartment. For therapeutic purposes, the low concentrations reflected in the terminal slope after IV injection may not have any practical relevance. As seen from Fig. 1, the relevant plasma concentrations persisted for only approximately 3 h after each dose.

Figure 1.

 Butorphanol concentrations in horses after IM (triangle) or IV (circle) administration at 0.08 mg/kg. Concentrations (ng/mL) represent the mean ± SE of six horses. Top panel is a shown as a logarithmic y axis and bottom panel is a linear y axis.

Differences in the terminal slope half-life can be misleading in a compartmental analysis. The plasma concentration profile (Fig. 1) shows that after peak concentration is attained, the two curves parallel each other closely indicating that clearance is probably not different from IV vs. IM dosing. Further evidence of this is that the clearance reported for the IM injection was reported per F (CL/F). The mean value of F was 0.37. Correcting the clearance value for F, yields a value of 276 mL/kg/h, (741.3 mL/kg/h × 0.37 = 276.2 mL/kg/h), which is almost identical to mean IV clearance (274.5 mL/kg/h).

The volume of distribution for butorphanol was approximately the same in the two studies (1.1 L/kg and 1.03 L/kg, respectively), but calculated clearance was lower in the current study (4.5 mL/kg/min and 21 mL/kg/min, respectively). The reason for the difference in clearance can once again be accounted for by the sensitivity of the assay. Because clearance is inversely proportional to the AUC (CL = Dose/AUC), and total AUC was higher in this study because of the longer terminal slope after IV injection, calculated clearance was lower. Clearance is a calculated pharmacokinetic parameter but does not imply that there was a physiologic difference in clearance mechanisms between the two studies.

An explanation for the low systemic absorption of butorphanol after intramuscular injection is not known. Absorption after IM injection in other species is quite high, often approaching 100% (Gaver et al., 1980; Pfeffer et al., 1980; Carroll et al., 2001; Riggs et al., 2008). Examination of the plasma profile (Fig. 1) shows that although absorption rate was fast, the peak concentration was very low suggesting a problem with drug loss from the injection site. A rapid absorption rate does not necessarily imply a high extent of systemic absorption because the absorption rate parameter (Ka) actually represents the disappearance of drug from the site of injection, rather than the appearance of drug in the systemic circulation (Toutain & Bousquet-Mélou, 2004). It suggests that there may be a fraction of the drug – perhaps 63%– which is not bioavailable.

To maintain targeted plasma concentrations above 10 ng/mL (the level suggested by previous studies) frequent administration is needed as suggested by Fig. 1. At this dose, injections every 3 h may be necessary. Because such frequent injections may be impractical in many clinical settings, a constant-rate infusion (CRI) that we reported previously may be a logical alternative to maintaining a persistent level of analgesia (Sellon et al., 2004). A CRI will also avoid high peak plasma concentrations that occur in the first 30 min after bolus IV injection that may be associated with adverse effects, and the potential myositis caused by repeated IM injections. The revised systemic clearance values for butorphanol, as determined in this study, suggest that a CRI of <13 μg/kg/hr would be effective at maintaining therapeutic concentrations of butorphanol in adult horses. However, additional study is needed to confirm the appropriate range of CRI dose that is likely to be clinically effective in horses.