Materials and Methods

Study design—The Colorado State University Animal Care and Use Committee approved this protocol. This study was performed in 2 phases over a period of 2 years. The first phase of the study was to determine the pharmacokinetics of morphine when given as morphine sulfate IV or IM to healthy adult llamas. The second phase of the study was to determine the pharmacokinetics and pharmacodynamics of 3 doses of morphine when given IV to healthy adult llamas. Data from both phases were combined for the pharmacokinetic analysis. All treatments were administered as a 1-time injection. Plasma and serum concentrations of morphine and M6G, an active morphine metabolite, were measured over a 24-hour period. Morphine-3glucuronide was not quantified.

Phase 1 experimental procedure—Six healthy llamas (3 sexually intact females and 3 neutered males) that were 5.8 ± 5.1 (mean ± SD) years old and weighed 144.0 ± 32.4 kg were studied. Llamas were housed in individual stalls in a climate-controlled barn. They were fed grass hay and water ad libitum.

Three treatments were administered to each llama in random order with a 1-week washout period between treatments. The morphine doses were prepared and administered from commercially available aqueous morphine sulfate^a (15 mg/mL), and the doses listed are based on the morphine sulfate quantity. The actual doses of morphine base are provided. The 3 treatments were as follows: IV administration of morphine sulfate at 0.05 mg/kg, (morphine base, 0.04 mg/kg) with IM administration of physiologic saline (0.9% NaCl) solution; IM administration morphine sulfate at 0.5 mg/kg (morphine base, 0.38 mg/kg) with IV administration of saline solution; and IV and IM administration of saline solution as a placebo. The volume of saline solution administered was matched with the volume of morphine given by the corresponding route. Individuals observing and assessing the llamas were blinded with respect to treatments.

Llamas were restrained in a chute^b for IV catheter placement. Skin over both jugular veins was clipped and aseptically prepared. One to 2 mL of 2% lidocaine hydrochloride^c was infiltrated into the skin over the desired location for IV catheter placement in the right jugular vein. A 1-cm-long incision was made through the skin. A 16-gauge, 20.3-cm-long IV catheter^d was placed into the right jugular vein and secured to the skin with 2-0 nylon suture. Llamas were restrained securely within their stalls for treatment injections. Treatments were administered IV into the left jugular vein by direct venipuncture. Treatments were administered IM into the right triceps muscles. Llamas were then loosely restrained within their stalls with access to hay and water.

Blood sample collection was performed in the following manner: 10 mL of blood was collected from the IV catheter into a syringe to clear any heparinized saline solution from the catheter, a separate 10-mL blood sample was collected with a different syringe and injected into a clean glass tube, the original 10 mL was injected back into the llama through the catheter, and then 10 mL of heparinized saline solution was rinsed through the catheter to prevent clotting. Blood was collected at baseline (time 0) and at 2, 5, 10, 15, 30, and 45 minutes and 1, 1.5, 2, 3, 4, 6, 8, 16, and 24 hours after treatment. Blood samples were refrigerated and centrifuged within 1 to 8 hours of collection. Serum was separated and frozen at −20°C until shipment to the laboratory^e for analysis.

Phase 2 experimental procedure—Six healthy sexually intact female llamas that were 10.7 ± 4.2 years old and weighed 166.1 ± 30.0 kg were studied. Three of the llamas used for year 2 were the same 3 adult females used in year 1 of the study. Because of the 1-year time difference, data from these llamas are treated as independent observations for purposes of pharmacokinetic and statistical analyses. Llamas were housed in individual stalls in a climate-controlled barn. They were fed grass hay and water ad libitum.

Three treatments were administered to each llama in random order with a 1-week washout period between treatments. The morphine doses were prepared as previously stated. The 3 treatments were as follows: IV administration of morphine sulfate^a at 0.05 mg/kg (morphine base, 0.04 mg/kg), IV administration of morphine sulfate at 0.25 mg/kg (morphine base, 0.19 mg/kg), and IV administration of morphine sulfate at 0.5 mg/kg (morphine base, 0.38 mg/kg). All doses of morphine sulfate were combined with saline solution by a pharmacist to create a total volume of 10 mL. Individuals observing and assessing the llamas were blinded with respect to treatments.

Catheter placement and llama restraint during data collection were the same as those for phase 1. For treatment administration, llamas were restrained in a chute^b while a second temporary IV catheter^f (20 gauge, 4.8 cm-long) was placed in the left jugular vein. An extension set was attached to the catheter during morphine sulfate administration. Morphine sulfate was injected over a 30-second period, the extension set and catheter were rinsed with 10 mL of saline solution, and then the catheter was removed.

Blood samples were collected in the same manner as in phase 1 except that heparinized vacuum tubes were used. Samples were refrigerated and centrifuged within 1 to 8 hours of collection. Plasma was separated and stored at −70°C until shipment to the laboratory^e for analysis.

Heart rate and respiratory rate were recorded once prior to treatment (baseline) and at 15-minute intervals for the first hour, then once each at 1.5, 2, 3, 4, 6, 8, 16, and 24 hours after treatment. Heart rate was determined by cardiac auscultation, and respiratory rate was determined by observation of thoracic excursions. Rectal temperature was recorded once prior to treatment (baseline), then at 30-minute intervals for 2 hours, and then once each at 3 and 4 hours after treatment. For any llamas with a rectal temperature > 38°C, their rectal temperature was checked hourly until it returned to < 38°C.

Sedation was recorded once prior to treatment (baseline), once at 5 minutes after treatment, and then at 15-minute intervals after treatment until return to baseline. Degree of sedation was evaluated and graded on a scale of 0 to 4, similar to that described by Grubb et al,¹ as follows: grade 0, no sedation; grade 1, mild sedation evidenced only by slight lowering of the head carriage or ear position or protrusion of the lower lip, delayed reaction to surroundings, or extension of the head and neck such that the weight of the head and neck was centered more over the shoulders than out in front of the body; grade 2, obvious sedation evidenced by signs such as those described for grade 1, prolapsed third eyelids, decreased awareness, and mild ataxia without recumbency; grade 3, obvious sedation evidenced by signs such as those described for grade 2 and recumbency with rising easily achieved by stimulation; and grade 4, obvious sedation evidenced by signs such as those described for grade 2 and recumbency without rising after stimulation. All sedation scores were assessed by 1 observer (SRU).

Analgesia was assessed with cutaneous electric stimulation prior to treatment (baseline), 5 minutes after treatment, and then at 15-minute intervals after treatment until return to baseline. Three consecutive and consistent voltage measurements were required to establish the baseline measurement and to confirm the return to baseline. Two 5 × 5-cm areas were clipped on the lateral aspect of the left forelimb at the level of distal portions of the MC and BR. Two adhesive electrode pads^g (2.3 × 4.6-cm adhesive with approx 1-cm diameter circular electrode contact surface in the center) were placed adjacent to each other at each location so that the widest edges of the adhesive pads were touching. This placement created a standard distance of approximately 2.3 cm between the centers of each electrode. Ultrasound gel was applied to the electrode on each pad prior to placement to maintain moisture and consistent contact throughout the testing period. Conductive cables were attached to each set of electrodes and labeled to identify the testing location (MC or BR). Cables were connected to a voltage stimulator^h with an in-line switch that controlled the application of the stimulus. The voltage stimulator was used to control the frequency, delay, duration, and volts applied to the electrodes. Frequency of the stimulus was 5 cycles/s, delay was 1 millisecond, and duration of the pulse was 10 milliseconds. Volts applied were variable depending on the response of each llama, but all testing started at 1.5 V. Voltage was increased in increments of 2 V until 10 V was reached, then voltage was increased in increments of 5 V until a response was evoked. A minimum interval of 30 seconds was allowed prior to restimulating the llamas when voltage was increased. The stimulus was terminated immediately when the llama responded with a deliberate limb lift or after 15 seconds if the llama did not respond. All analgesic assessments were performed by 1 observer (SRU).

Plasma and serum morphine concentrations—Plasma and serum morphine concentrations were determined by high-pressure liquid chromatography with electrochemical coulometric detection, as described elsewhere.²⁶ Analyses of serum samples were compared with analyses of plasma samples to validate the assay in either matrix for the high-pressure liquid chromatography analysis of morphine and M6G. The sample matrix was fortified with morphine standard solutions to prepare calibration standards. The standard curves were accepted if the R² value (coefficient of determination) was ≥ 0.99 and the measured concentrations were within 15% of actual concentrations. Interday accuracy and precision were not measured. A new calibration curve was not used for analysis on each day. The limit of quantification was 5 ng/mL for morphine and 25 ng/mL for M6G.

Pharmacokinetic analysis—Pharmacokinetic analysis was performed with a computer program.ⁱ Noncompartmental analysis was performed with a 1/ŷ² weighting factor, where ŷ was the actual plasma concentration and the AUC_0–∞ was calculated by the linear trapezoidal rule. Compartmental analysis was performed with a 1/ weighting factor, where was the predicted plasma concentration. The best-fit compartmental model was assessed by residuals and the Akaike information criterion.²⁷ The bioavailability of morphine following IM administration was determined with the noncompartmental parameters by dividing the AUC_0–∞ for IM administration of 0.5 mg/kg by the AUC_0–∞ for IV administration of 0.5 mg/kg. Linearity among doses administered IV was determined by dividing the AUC_0–∞ and C₀ by dose to normalize these parameters.

Pharmacokinetic parameters for morphine were also evaluated for dogs and cats.^9,10 We are unaware of any pharmacokinetic studies of morphine in any ruminant species published to date. The studies^9,10 on dogs and cats cited here were consistent in the assay method used for measurement of morphine and its metabolites; these methods were specific for morphine and were capable of differentiating morphine from its major metabolites. Studies^7,8,13 have been performed in which morphine concentrations in other species were determined to evaluate different doses or routes of administration, but these studies used nonspecific techniques (eg, immunoassays) that may not be useful in differentiating between morphine and its metabolites.

Pharmacodynamics—Pharmacodynamics were assessed by fitting the plasma concentration-versus-response relationship to a curve to compare the independent variable of plasma concentration with the response variables of analgesia, sedation, and any physiologic parameters that differed significantly among the 3 doses of morphine administered IV.

Statistical analysis—This study had a randomized crossover design. Statistical analyses were performed by use of a computer program.^j,k Descriptive values for the pharmacokinetic parameters are reported as mean ± SD. For comparisons of the pharmacokinetic parameters among doses and between routes of administration, an ANOVA was performed.^j Natural logarithmic values were analyzed by use of the ANOVA to normalize variance, but actual values are presented in the text. The effects analyzed were morphine sulfate dose (fixed) and llama (random). Least squares means (adjusted means) were compared with pairwise t tests for dose response. For comparison between phases 1 and 2 of the groups treated IV with morphine at 0.05 mg/kg, an ANOVA was performed^k and the fixed effect of year was analyzed. Values of P < 0.05 were considered significant.

A repeated-measures ANOVA was performed on all physiologic, sedation, and analgesic data.^j The following fixed effects were analyzed: treatment, time, and interaction of treatment × time. Random effects were llama and the interactions of llama with time and treatment. Pairwise t tests were performed on adjusted means of treatment differences at each time point for response variables that had an overall significance with a value of P < 0.05. A value of P < 0.05 was considered significant for t test results. A period effect was analyzed with week as a fixed effect, but it was removed from the model because it was not significant for any response variable. Autoregressive error terms (subject = llama × treatment interaction) were considered for each response for possible changes over time, and they were found to be significant (P ≤ 0.01) in all cases. All voltage measurements were converted to natural logarithmic values, and the difference from baseline was the parameter of interest for the ANOVA. Actual values are presented in the text. Voltage results for the MC and BR regions were analyzed separately.

Peak sedation score, time to reach peak sedation, and time to return to baseline sedation were analyzed with a randomized complete block model.^k Treatment was analyzed as a fixed effect, and llama was analyzed as a random effect. A Ryan-Einot-Gabriel-Welsch multiple range test was performed on the means of each response variable to control type I experiment-wise error rate. Pairwise t tests were performed on the adjusted means. A value of P < 0.05 was considered significant. Studentized residuals were plotted by llama and treatment to examine variance for each parameter. Variance was stable for sedation parameters; therefore, no transformation of data was required.

A randomized complete block analysis was similarly performed for peak voltage, time to reach peak voltage, and time to return to baseline voltage. Voltage results for the MC and BR regions were analyzed separately. For both regions, individual differences from baseline of natural logarithmic values of peak voltage were analyzed to stabilize variance. Time to reach peak voltage for the BR region required a logarithmic transformation to stabilize the variance; no other parameters required variance stabilization. Actual numbers are presented in the text.

Results

Pharmacokinetics—No significant differences were found among any parameters between phases 1 and 2 of the groups treated IV with morphine at 0.05 mg/kg; therefore, data from these groups were pooled for the descriptive pharmacokinetic statistics. Data from 1 llama were removed from the analysis because morphine was detectable at only 1 time point. Pharmacokinetic variables for IV administration of low-dose (0.05 mg/kg) morphine sulfate were determined (Table 1). Low-dose morphine sulfate administration resulted in a rapid decrease in morphine plasma concentrations, which decreased to undetectable concentrations within 15 minutes in all llamas representing only the distribution phase of morphine (Figure 1). Because of the sparse time points for this dose, a noncompartmental analysis was used. Because so few data points were available for analysis and the likelihood that a true terminal slope was not detected, the pharmacokinetic variables were considered unreliable for comparison to the other morphine doses. These data are summarized here only for completeness and will not be used for further comparisons.

View Full Size

Semilogarithmic graph of mean ± SD plasma concentrations of morphine sulfate following IV administration at 0.05 mg/kg to 11 healthy llamas. Data for experimental phases 1 and 2 are pooled.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

No significant differences were found between plasma and serum analyses for morphine over the 2 phases of the study; therefore, results were combined for the pharmacokinetic analysis. Following IV administration of morphine sulfate at 0.25 and 0.5 mg/kg, plasma profiles were biphasic with a rapid distribution phase followed by a slower elimination phase to which a 2compartment open model best fit. Normalized ratios for AUC_0–∞ and C₀ demonstrate dose linearity between IV administration of morphine sulfate at 0.25 and 0.5 mg/kg in llamas. For 0.25 and 0.5 mg/kg, respectively, the AUC_0–∞/dose was 545 ± 225 and 484 ± 133 and the C₀/dose was 4,431 ± 2,518 and 4,141 ± 2,957.

Pharmacokinetic parameters for the 2-compartment analyses of IV administration of medium-dose (0.25 mg/kg) and high-dose (0.5 mg/kg) morphine sulfate were determined (Table 2). The AUC_0–∞ was significantly (P < 0.001) different between treatments; all other parameters were not significantly different. Plasma concentrations of the medium and high doses were similar over time (Figure 2).

View Full Size

Semilogarithmic graph of mean ± SD plasma concentrations of morphine sulfate following IV administration at 0.25 mg/kg (closed circles) and 0.5 mg/kg (open circles) to 6 healthy llamas. Time 0 represents the time of morphine sulfate administration.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Pharmacokinetic variables for IM administration of morphine (0.5 mg/kg) were determined (Table 2; Figure 3). A 2-compartment open model best described the plasma profile with an absorption phase for 5 of 6 llamas. Data from 1 llama best fit a 1-compartment model and were excluded from compartmental analysis for IM administration.

View Full Size

Semilogarithmic graph of mean ± SD plasma concentrations of morphine sulfate following IM administration at 0.5 mg/kg to 5 healthy llamas. Time 0 represents the time of morphine sulfate administration.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Bioavailability of morphine sulfate following IM administration at 0.5 mg/kg was 120 ± 30%. Bioavailability was determined from noncompartmental analyses of IM and IV administration of morphine at 0.5 mg/ kg. The remainder of parameters reported here is from the compartmental analyses of IM and IV administration of morphine at 0.5 mg/kg. The distribution halflife was significantly (P < 0.001) higher for the IM route than for the IV route (0.5 ± 0.08 hours vs 0.03 ± 0.02 hours, respectively). The y-axis intercept for the distribution phase was significantly (P < 0.001) lower for the IM route than that for the IV route (139 ± 33 ng/mL vs 1,948 ± 1,412 ng/mL, respectively). The y-axis intercept for the elimination phase was also significantly (P = 0.035) lower for the IM route than that for the IV route (27 ± 11 ng/mL vs 123 ± 80 ng/mL, respectively). No other parameters of interest were significantly different between IM and IV administration (Table 2).

Morphine-6-glucuronide was detected infrequently and only at concentrations of < 40 ng/mL following IV and IM administration of morphine (0.5 mg/kg). These results were inconsistent and are not reported. No peaks were found in the chromatogram that coincided with morphine-3-glucuronide, but nothing further was done to quantitate its presence.

Pharmacokinetic parameters for morphine were compared among llamas, cats, and dogs (Table 3). Because our assay method was specific for morphine and its major metabolites, results from our study are only comparable to other studies in which similar methods were used.

Behavior observations—Five of 6 llamas receiving high-dose morphine had muscle tremors for approximately 1 hour after treatment. None of the llamas receiving the low and medium doses had muscle tremors. One llama became hyperexcitable and extremely sensitive to movement and sound within 10 minutes after receiving high-dose morphine. This llama concurrently had signs of sedation such as drooping ears, eyelids, and lips; ataxia; and altered head and neck carriage. This behavior did not affect the overall analysis of peak sedation because this llama had more typical sedation behavior prior to the onset of hyperexcitability.

Physiologic parameters—The main effect of interest was the interaction of treatment × time. The interaction of treatment × time for heart rate did not differ significantly (P = 0.307) among the 3 treatment groups. The interaction of treatment × time for respiratory rate also did not differ significantly (P = 0.424) among treatment groups; however, treatment alone was significant (P = 0.031) overall. All doses of morphine caused a decrease in respiratory rate (Figure 4). The low-dose morphine group was minimally affected, whereas respiratory rate of the medium- and high-dose morphine groups had a greater and similar decrease. None of the groups returned to baseline values by 24 hours after treatment.

View Full Size

Mean respiratory rates over time following IV administration of morphine sulfate at 0.05 mg/kg (circles), 0.25 mg/kg (squares), and 0.5 mg/kg (triangles) to 6 healthy llamas. This graph represents results up to the first 8 hours after treatment. Time 0 represents the time of morphine sulfate administration. _a,bDifferent letters at a given time point indicate values that are significantly (P < 0.05) different from each other.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

The interaction of treatment × time for body temperature was significant (P = 0.001) overall. Mean body temperature increased as the morphine dose increased (Figure 5). Little change in body temperature occurred in the low-dose morphine group. Peak mean body temperature in the medium- and high-dose morphine groups occurred between 90 to 120 minutes after treatment. Mean body temperatures in the medium-dose morphine group returned close to baseline by 4 hours after treatment. Mean body temperatures of the highdose morphine group did not return to baseline by 4 hours after treatment. Three of 6 llamas in the high-dose morphine group had body temperatures that remained above baseline until 6 to 8 hours after treatment.

View Full Size

Mean rectal temperatures over time following IV administration of morphine sulfate at 0.05 mg/kg (circles), 0.25 mg/kg (squares), and 0.5 mg/kg (triangles) to 6 healthy llamas. Time 0 represents the time of morphine sulfate administration. _a–cDifferent letters at a given time point indicate values that are significantly (P < 0.05) different from each other.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Sedation score—Sedation was present for ≤ 2 hours for any dose (Figure 6). The treatment × time interaction for sedation score was significant (P < 0.001) overall. Peak quality of sedation was significantly (P = 0.002) different among the treatments, with higher doses producing greater sedation. Time to reach peak sedation did not differ significantly (P = 0.114) among the groups; however, time to return to baseline did differ significantly (P = 0.002), with higher doses producing longer-lasting sedation.

View Full Size

Mean sedation scores over time following IV administration of morphine sulfate at 0.05 mg/kg (circles), 0.25 mg/kg (squares), and 0.5 mg/kg (triangles) to 6 healthy llamas. Sedation was scored on a scale of 0 to 4 (See Materials and Methods section for details). Time 0 represents the time of morphine sulfate administration. See Figure 5 for remainder of key.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Analgesia—Overall, no significant differences were found for voltage applied to either the MC (interaction of treatment × time; P = 0.526) or BR (interaction of treatment × time; P = 0.337) regions (Figure 7). Responses to electric stimulation were highly variable among individuals at baseline and throughout the testing periods for each treatment. Individual peak voltage measurements (calculated as differences from the baseline value of each individual) were determined for each llama at the BR region for the 3 morphine doses (Figure 8). The term peak can refer to either a negative or positive value because some llamas had a reduced tolerance to electric stimulation after morphine treatment, whereas other llamas had greater or no change in tolerance. Testing at the MC region yielded results that were more highly variable than those for the BR region.

View Full Size

Mean ± SD peak voltage differences from baseline, as a measure of analgesia, in response to electric stimulation of the distal portions of the MC (open bars) and BR (closed bars) following IV administration of morphine sulfate at 0.05, 0.25, and 0.5 mg/kg to 6 healthy llamas. A negative value indicates decreased tolerance to electric stimulation.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Individual peak voltage differences from baseline, as a measure of analgesia, in response to electric stimulation of the distal portion of the BR following IV administration of morphine sulfate at 0.05, 0.25, and 0.5 mg/kg to 6 healthy llamas. A negative value indicates decreased tolerance to electric stimulation.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Variation in patterns of peak voltage responses in llamas was not consistent among the 3 doses at each location. However, for the BR region, 4 of 6 llamas had an increased peak tolerance to electric stimulation (5 to 7 V) when given morphine IV at 0.25 mg/kg, compared with 3 llamas having an increased peak tolerance (5 to 20 V) when given morphine IV at 0.5 mg/kg; no llamas had an increased peak tolerance when given morphine IV at 0.05 mg/kg. However, the randomized complete block results for peak voltage responses were not significant (MC region, P = 0.637; BR region, P = 0.653; Table 4).

Time to reach a peak voltage response did not differ significantly (P = 0.265) among treatments for the MC region (Table 3). For the BR region, the low-dose morphine group reached a peak voltage response significantly (P = 0.041) more quickly than the medium-dose morphine group (Table 4).

Time to return to baseline voltage response did not differ significantly among treatments for the MC (P = 0.121) or BR (P = 0.408) region (Table 4). Return of voltage responses to baseline was difficult to assess. In the low-dose morphine group, 2 llamas had voltage responses that did not return to baseline at the MC region; 1 response remained 5 V below baseline, and 1 remained 2 V above baseline. Also, in the lowdose morphine group, 1 llama had a voltage response that remained 10 V below baseline at the BR region. In the medium-dose morphine group, 2 llamas had voltage responses that did not return to baseline at the BR region; 1 response remained 5 V below baseline, and 1 remained 7 V above baseline. In the high-dose morphine group, 2 llamas had voltage responses that remained below baseline at the MC region (5 and 10 V, respectively), and 1 llama had a voltage response that remained below baseline at the BR region (4 V). Because some analgesic responses did not return to baseline, time to return to baseline could not be determined.

Pharmacodynamic modeling—Response variables analyzed for pharmacodynamic modeling were respiratory rate, temperature, sedation score, and voltage responses for MC and BR regions. Although voltage responses for the MC and BR regions were not significant response variables, they were considered a priori parameters of interest for the pharmacodynamic modeling. The only response variable with a significant relationship to plasma concentration was sedation, which had a sigmoidal relationship (Figure 9).

View Full Size

Comparison of sedation score versus plasma concentrations of morphine for 6 healthy llamas by use of the following equation for a sigmoidal E_max model: E = (E_max × C^γ)/(C^γ + EC^γ₅₀). In this model, E_max = 1.520359, EC₅₀ = 85.434443, and γ = 2.773487. Plasma concentrations following IV administration of morphine at 0.25 mg/kg were plotted against the sedation score (circles indicate observed values, and line indicates predicted values). Sedation was scored on a scale of 0 to 4 (See Materials and Methods section for details). E_max = Maximal response. E = Effect. C = Plasma concentration. EC₅₀ = Plasma concentration that produces a 50% maximal response. γ= Shape parameter.

Citation: American Journal of Veterinary Research 68, 1; 10.2460/ajvr.68.1.25

• Download Figure
• Download figure as PowerPoint slide

Discussion

In our study, pharmacokinetics of morphine sulfate were characterized for 3 doses administered IV (0.05, 0.25, and 0.5 mg/kg) and 1 dose administered IM (0.5 mg/kg) to 6 healthy adult llamas. Considerable variation of pharmacokinetic parameters was found among individuals for all doses and routes of morphine sulfate administration. Pharmacodynamics of morphine sulfate were characterized for 3 doses administered IV (0.05, 0.25, and 0.5 mg/kg) to 6 healthy adult llamas. The response variables measured included heart rate, respiratory rate, body temperature, sedation, and analgesia.

Total body clearance was similar between IV administrations of medium- and high-dose morphine sulfate. The AUC_0–∞ and C₀ were much lower for IV administration of medium-dose morphine sulfate, compared with high-dose morphine sulfate, as expected. When these parameters were normalized for dose, they were similar, which indicates that linearity exists between IV administrations of these doses. Pharmacokinetic linearity implies that clearance is not dependent on the dose administered.²⁸ Total clearance approached hepatic plasma flow. Estimated cardiac output for animals that weigh 144 kg is approximately 70 mL/kg/min, determined on the basis of allometric scaling.²⁹ Assuming that hepatic blood flow is approximately 30% of cardiac output, maximum hepatic plasma clearance is 21 mL/kg/min. In the llamas of our study, total clearance was only slightly higher than this value, but this assumes equal partitioning of morphine in blood and plasma. We can anticipate some partitioning into blood because of the high Vd_ss observed, which causes plasma clearance to overestimate blood clearance. Therefore, it is reasonable to conclude that morphine in llamas can probably be categorized as a high-clearance (flow-dependent) drug. High hepatic clearance implies that the elimination is strongly affected by changes in hepatic blood flow and less affected by changes in intrinsic hepatic clearance (eg, changes in hepatic enzyme activity) and protein binding.

The Vd_ss appeared to be higher for IV administration of high-dose morphine sulfate than for medium-dose morphine sulfate, but this difference was not significant because of large variances in the volumes of distribution. For both doses, the mean Vd_ss exceeded unity, which suggests drug sequestration out of the sampling compartment. The mean terminal half-life was longer following IV administration of high-dose morphine sulfate than for medium-dose morphine sulfate, but this is a reflection of the dependence of half-life on volume of distribution and clearance. Because clearance remained relatively constant but Vd_ss increased, the calculated half-life increased proportionately.

Total body clearance following IV administration of high-dose morphine sulfate was similar to the clearance per bioavailability following IM administration of the same dose. The AUC_0–∞ was also similar between the 2 routes of administration. This similarity in AUC_0–∞ supports the observation that absorption of the dose administered IM was complete. However, absorption also was delayed and morphine was detectable in plasma approximately twice as long for the IM route (approx 8 hours) as for the IV route (approx 4 hours).

In comparing certain pharmacokinetic parameters of llamas to those of cats and dogs, we found that most of the llama parameters are similar to those of cats. The exceptions are terminal half-life and mean residence time of IM morphine administration, both of which are higher for llamas than cats. Terminal half-life of morphine administered IV and bioavailability of morphine administered IM were the only similar parameters between llamas and dogs.

Except for 1 report³⁰ of ponies that developed an increased heart rate, most species develop bradycardia subsequent to morphine administration.⁴ Bradycardia is a secondary effect of vagal stimulation that follows an initial increase in heart rate following morphine administration.⁴ Heart rate in llamas appeared to be unaffected by IV morphine administration at the doses tested in our study.

As with most species, llamas had a change in respiratory rate subsequent to morphine administration. Ponies develop increased respiratory rates for up to 4 hours following morphine administration.³⁰ However, many species develop respiratory depression that manifests as a decreased response to carbon dioxide in the blood.⁴ Results of our study suggest that IV administration of increasing doses of morphine caused a decreased respiratory rate in llamas. Further study of IV administration of a greater range of doses is required to determine whether a dose-response relationship exists for the respiratory effects of morphine. Arterial partial pressures of oxygen and carbon dioxide were not measured in our study; therefore, it is unclear whether the decreased respiratory rate that developed in these llamas reflected true respiratory depression. However, at no time did any of the llamas appear to be compromised, on the basis of alertness and mucous membrane color, as a result of their decreased respiratory rate. Although respiratory rates for all 3 groups did not return to baseline, it is possible that baseline values were increased as a result of the llamas being excited at the beginning of data collection periods. Respiratory rates measured ≥ 8 hours after treatment may be more representative of typical rates for llamas. However, the analysis was repeated without the baseline values, and significant differences between groups remained.

The effect on rectal temperature in our study was unexpected. Morphine can alter thermoregulatory centers in the hypothalamus when administered systemically.^21–25 Cats develop continually increasing body temperatures with increasing doses of morphine regardless of muscle activity.^21,25 Rats experience a ceiling effect such that their body temperature increases with increasing doses of morphine up to IP administration of 10 mg/kg²² or SC administration of 32 mg/kg.²³ Beyond these doses, rats develop hypothermia. In mice, changes in body temperature are dependent on ambient temperature regardless of morphine dose.²⁴ Results of our study suggest that llamas respond to morphine similarly as do cats and rats; as morphine dose increases, body temperature increases.^21–23,25

It is unclear whether the increase in rectal temperature is also caused by muscle tremors that develop following administration of high-dose morphine. In our study, muscle tremors occurred only in the highdose morphine group; therefore, is it likely that the increase in body temperature is associated primarily with thermoregulatory alteration and not muscle tremors. Further study of IV administration of a wider range of doses is required to determine whether a dose-response relationship of morphine on body temperature exists.

Results of our study suggest that llamas had a higher degree of sedation as morphine dose increased. However, at least 1 llama had some signs of hyperexcitability in the high-dose morphine group. Central nervous system depression at low doses and excitement at higher doses occur in most species; however, cats and horses appear to be more sensitive to the excitatory effects of opioids.^4,6,15 Llamas may be similar to cats and horses in this manner. However, hyperexcitability may not develop in the presence of severe or sustained pain, as in clinical situations for which analgesic treatment is indicated. Further study of IV administration of a higher range of morphine doses as well as clinical studies on llamas with signs of pain will be required to understand this effect more clearly.

Analgesia in llamas is difficult to consistently assess. In preliminary trials for our study, 3 methods of analgesia testing were compared for ease of use and consistency of results: thermal, mechanical, and electric. Cutaneous electric stimulation provided consistent, reproducible results for multiple llamas and for up to 6 hours of intermittent testing. Subcutaneous electric stimulation has been used to test the analgesic effects of buprenorphine, methadone, flunixin meglumine, and xylazine in sheep.³¹ An electric stimulus may not be the optimal test for an opioid because this type of stimulus is peripheral and acute, whereas morphine is more effective in altering slow fiber pain sensations rather than fast fiber pain sensations.³¹ However, electric stimulation provided the most consistent results of any analgesic test attempted on llamas in our study.

Despite the consistency of test results prior to the study and at baseline for individual llamas, reliable analgesia could not be demonstrated for any of the doses of morphine administered to any of the llamas. The 2 sites tested were selected to represent areas with more or less muscle tissue for analgesic testing. By selecting these 2 areas, results of our study may have been used to detect differences between superficial and deep analgesic effects of morphine as represented by the MC and BR sites, respectively. This difference in muscle tissue may have contributed to the variable patterns of analgesia between the 2 sites. However, because of high individual variability among llamas, no consistent patterns could be found for analgesia at either site. Although overall results for individual llamas were highly variable, medium-dose morphine provided the most consistent increase in tolerance to electric stimulation at the BR region; consistency of these results suggests that medium-dose morphine may provide some deep analgesia.

In conclusion, IV and IM administration of morphine to healthy llamas resulted in plasma concentrations that are in a similar range as those observed in other species. A large apparent volume of distribution and high systemic clearance characterized morphine administration in llamas. Intramuscular injection prolonged the terminal half-life but did not affect the extent of the systemic availability. The clinical relevance of plasma concentrations produced in our study in relation to physiologic and behavioral characteristics will require further exploration.

On the basis of pharmacokinetic results and physiologic findings in our study, a suggested dose and dosing interval for additional studies are 0.25 mg/kg administered IV every 4 hours. Minimal adverse effects were observed for this morphine dose and route among the llamas in our study, and it provided the most consistent deep analgesic effect. On the basis of physiologic parameters of respiratory rate and rectal temperature as well as the sedation score and pharmacokinetic results, a dosing interval of 4 hours should minimize the chances of unsafe accumulations of morphine in the body that could negatively affect the physical well-being of a llama. It is important when applying this information to consider that our study was performed on healthy llamas that did not have signs of pain. Morphine administration to llamas with signs of pain should be closely monitored and adjusted to meet the needs of the individual. Individual variation of response to morphine is high among llamas without signs of pain, and this variation is also likely to be high among llamas with signs of pain. Further studies of other morphine doses and routes of administration are needed to determine a safe dose and dosing interval for llamas. A more appropriate method is needed for testing analgesia in llamas. Further studies are also needed of clinical applications of morphine to llamas with clinical signs of various types of pain.