• Open Access

Pharmacokinetics and Safety of Single and Multiple Oral Doses of Meloxicam in Adult Horses


  • Preliminary findings have been presented at the Bain-Fallon Conference, Equine Veterinarians Australia, Maroochydore, Queensland, 2009

Corresponding author: S.L. Raidal, Department of Clinical Studies, New Bolton Center, University of Pennsylvania, Kennett Square, PA 19348; email: sraidal@csu.edu.au.



Safety of meloxicam, a potent NSAID with selective COX-2 inhibition, has not been evaluated in horses.


To evaluate pharmacokinetics and safety of single and repeated oral doses of meloxicam in adult horses.


Forty-nine healthy, university-owned adult lightbreed horses.


Study conducted in 2 parts. Part I addressed pharmacokinetics of single oral dose meloxicam (0.6 mg/kg) in 16 horses. Part II, 33 horses were randomly assigned to 5 treatment groups to assess prolonged administration (0.6 mg/kg PO q24h for 6 weeks, n = 7) or higher doses (1.8 mg/kg, n = 7, or 3.0 mg/kg PO q24h, n = 7) of meloxicam for 2 weeks, compared with control horses (placebo, n = 7, or phenylbutazone, 4.4 mg/kg q12h on day 1, 2.2 mg/kg q12h for 4 days, then 2.2 mg/kg q24h for 9 days, n = 5).


Maximum plasma concentration following a single oral dose of meloxicam was 915.1 ± 116.9 ng/mL and elimination half-life 10.2 ± 3.0 hours. Meloxicam (0.6 mg/kg, q24h, PO for 6 weeks) yielded plasma concentrations between 100 and 1000 ng/mL and was well tolerated by healthy adult horses. Administration of 3–5 times the recommended dose of meloxicam was associated with decreased total serum protein and albumin concentrations, gastrointestinal damage, renal damage, or bone marrow dyscrasia. PBZ administration was associated with the development right dorsal colitis, gastric ulceration, and protein losing enteropathy in 2 horses.

Conclusions and Clinical Importance

Administration meloxicam at 0.6 mg/kg q24h was well tolerated for 6 weeks, without drug accumulation in plasma. Higher doses were associated with dose-dependent adverse effects typical of class of drugs.




nonsteroidal anti-inflammatory drug




standard deviation


standard error of mean

Nonsteroidal anti-inflammatory drugs (NSAIDs) are commonly prescribed in equine practice as analgesics and anti-inflammatories. Such effects are mediated largely by reduced prostanoid production due to inhibition of the cyclooxygenase (COX) enzyme. Two COX isoenzymes, coded by separate genomic sequences, have been recognized since the mid-1990s.[1, 2] COX-1 is constitutively expressed in nearly all cells and was initially believed to be responsible for physiological beneficial effects associated with some prostaglandins, notably gastrointestinal and renal homeostatic mechanisms. In contrast, COX-2 is induced by proinflammatory mediators such as cytokines, mitogens, and endotoxin, and was believed to mediate detrimental effects associated with prostaglandins, such as pain, inflammation, and pyrexia. This dichotomous view of COX-1 and COX-2 isoenzymes has broken down somewhat with evidence that COX-2 is constitutively expressed in many tissues, notably central nervous system, kidney, bone, and joints.[3-5] NSAIDs are still, however, commonly assessed in terms of their selectivity for each isoenzyme, and the use of selective COX-2 agents might afford an increased therapeutic index by sparing constitutive COX-1 activity to maintain homeostatic functions.[6]

Most NSAIDs routinely used in equine practice are older, relatively nonselective COX inhibitors which target COX-1 and COX-2 in equal measure. Adverse effects associated with the administration of nonselective COX inhibitors have been well documented in adult horses and include right dorsal colitis,[7-9] gastrointestinal ulceration,[10-12] and nephrotoxicity.[13, 14] Meloxicam is a potent NSAID that is an effective analgesic and anti-inflammatory drug in adult horses at a dose rate of 0.6 mg/kg,[15-17] with preferential inhibition of the cyclooxgenase enzyme (COX-2).[18, 19] The current study was conducted in 2 parts designed to investigate the pharmacokinetics of single and repeated doses of an oral suspension of meloxicam, and to evaluate the safety of repeated doses of this medication relative to the safety of an appropriate control and the nonselective COX-inhibitor phenylbutazone (PBZ).

Materials and Methods

Experimental Design and Research Horses

This study was conducted in 2 separate parts and all procedures were approved by the Animal Care and Ethics Committee at Charles Sturt University (approval numbers 07/064 and 07/127). All horses were teaching or research horses resident at Charles Sturt University. Only horses older than 2 years and with hematology, serum biochemistry, and physical examination parameters within acceptable limits were eligible for inclusion in the study. Single dose pharmacokinetics in fed horses were determined in Part I using 16 healthy light horses (Thoroughbreds or Standardbreds, 10 geldings and 6 mares) with mean bodyweight 522.3 ± 60.8 kg (SD), aged 3–19 years. Horses were enrolled in the study in 2 groups of 8 and each horse received a single oral dose equivalent to 0.6 mg/kg of meloxicam.1 Plasma samples were collected before administration of meloxicam and at 1, 1.5, 2, 2.5, 3, 4, 5, 6, 8, 10, 12, 16, 24, 36, 48, and 72 hours after treatment, for determination of meloxicam concentrations,

Part II was conducted 12 months after Part I to examine the safety of meloxicam1 administered at 1.8 mg/kg PO q24h (3 times the recommended dose rate) and 3.0 mg/kg PO q24h (5 times the recommended dose rate) for a period of 14 days, as well as at 0.6 mg/kg PO q24h (the recommended dose rate) for 42 days (3 times the recommended duration of treatment). Treatment effects were compared with those observed in horses receiving placebo (product vehicle) only for 6 weeks and in horses receiving phenylbutazone (PBZ2 4.4 mg/kg q12h on first day, 2.2 mg/kg q12h for 4 days, and then 2.2 mg/kg q24h for 9 days). Thirty-three horses (16 mares and 17 geldings) with mean body weight of 508.7 ± 10.54 kg (SD) and mean age of 8.7 ± 4.8 years (SD, range; 3–20 years) were randomly assigned (by ballot) to 5 treatment groups: Treatment Group A (control, placebo3 administration for 6 weeks, n = 7), B (meloxicam 0.6 mg/kg for 6 weeks, n = 7), C (meloxicam 1.8 mg/kg for 2 weeks, n = 7), D (meloxicam 3.0 mg/kg for 2 weeks, n = 7), and E (PBZ 4.4 mg/kg q12h on first day, 2.2 mg/kg q12h for 4 days, then 2.2 mg/kg q24h for 9 days, n = 5). This study was conducted as 6 consecutive replicate groups over a 7-month period. Replicate Groups from 1 to 5 each contained 5 horses such that there was 1 horse receiving each treatment in each replicate group. Replicate Group 6 contained 2 horses each receiving treatment A, B, C, or D (8 horses). Personnel responsible for data analysis and veterinary examination of all horses were blinded to treatment. Analysis of variance demonstrated no significant difference between Treatment Groups in Part II with respect to age (P = .895) or body weight (P = .488).

Throughout Part II horses were examined twice daily and blood samples were collected via jugular catheters for determination of plasma concentrations of meloxicam from Treatment Groups A (control) and B (meloxicam 0.6 mg/kg PO q24h) on Day 0, Day 6, Day 13, and Day 41 at the following intervals: 0, 1, 1.5, 2, 3, 3.5, 4, 5, 6, 8, 12, and 24 hours after administration. Additional samples from Group A and B were collected on Days 1–14 by jugular venipuncture immediately before and 1.5 hours after treatment to determine plasma trough and peak drug concentrations, respectively. Drug excretion data were determined from samples collected on Days 42 to 45 at 24, 48, 72, and 96 hours after administration of the final drug dose. Samples from treatment Groups C (meloxicam 1.8 mg/kg PO q24h), D (meloxicam 3.0 mg/kg PO q24h), and E (PBZ) were obtained on Days 0, 6, and 13 at the following intervals: 0, 2, 4, 8, 12, and 24 hours after administration. Sham collections were performed at other times to preserve blinding. Additional samples were obtained from Group E (PBZ) horses at 48, 72, and 96 hours after administration of the final drug dose on Day 13.

Blood was collected twice weekly from all horses for routine hematology and serum biochemistry by a commercial laboratory.4 Gastroscopy, abdominal ultrasound, and urinalysis were performed on Day 0 and thereafter weekly throughout the duration of the study (Days 6 and 13, all horses; plus Days 20, 27, 34, and 41 for Groups A and B). Gastroscopy was performed using a 3 m endoscope5 after overnight fasting. Horses were sedated immediately before the procedure (xylazine 0.04 mg/kg and acetylpromazine 0.02 mg/kg by intravenous administration) and the severity of gastric ulceration was scored independently by 2 researchers blinded to treatment using a grading system adapted from previous publications.[20, 21] Briefly, the cardia, greater and lesser curvature, and margo plicatus were examined for ulceration of the squamous epithelium and graded 0 (intact epithelium with or without areas of reddening and hyperkeratosis), 1 (small single or multifocal ulcers), 2 (large single or multifocal lesions, or extensive superficial ulcers), or 3 (extensive, often coalescing ulcers with areas of deep ulceration). The scores from both researchers were summed for analysis. Hyperkeratosis of the squamous mucosa and edema or hyperemia of the glandular mucosa were noted, but not included in analyses. The presence and severity of lesions in the pyloric antrum were graded subjectively by a single examiner blinded to treatment (Grade 0, intact mucosal epithelium, uniform or relatively uniform color; Grade 1, obvious patchy erythema, multiple superficial erosions or both; Grade 2, extensive superficial erosions (often circumferential or linear), with or without areas of deeper ulceration).

Transabdominal ultrasound6 was performed after gastroscopy using convex array (8–1 MHz) or microconvex array (7.5–5 MHz) probes to assess renal parenchyma and right dorsal colon wall thickness at intercostal spaces (ICS) from 14 to 10, according to published techniques.[22-24] Urine was collected by free catch or urethral catheterization from every horse for urinalysis by reagent strips and urine enzyme analysis. Sediment examination was performed only on urine samples that returned a positive result for protein. Reagent strip analysis and sediment examination were performed at Charles Sturt University within 120 minutes of sample collection. Samples for urine enzyme analysis (GGT and ALP) were chilled (4°C) within 30 minutes of collection and dispatched overnight on ice to a commercial laboratory5 for analysis. Fecal samples were collected for detection of fecal occult blood using guiaic slides,7 as previously described.[25] Sternal bone marrow aspirates were collected for cytology on Day 0 and Day 13 (all horses), and Day 41 (Group A and B horses) as previously described.[26]

Sample Analysis

Meloxicam and PBZ concentrations in plasma were measured by ultra high performance liquid chromatography with ultraviolet detection, using piroxicam as an internal standard for meloxicam assays and after protein denaturing using acetonitrile. Meloxicam and phenylbutazone were extracted from 500 μL plasma. Chromatographic separations were performed using an Acquity UPLC8 fitted with a reverse-phase 100 mm C18 (octadecyl silane) column (1.7 μm particles; 100 × 2.1 mm)9 and a gradient mobile phase consisting of 0.1% trifluoroacetic acid in water and acetonitrile. The flow rate was 0.4 mL/min with a total run time of 6 minutes and an injection volume of 5 μL. The detection wavelength was 355 nm. For meloxicam, method recovery was >96%, calculated from a calibration curve derived from meloxicam standard peak area and piroxicam internal standard peak area. A linear relationship between detector response and plasma meloxicam concentrations from 20 to 4000 ng/mL was demonstrated graphically and using regression analysis. The method limit of quantitation was 20 ng/mL. The intrasample and intersample coefficients of variation were <3 and <10%, respectively. For phenylbutazone, the detection wavelength was 240 nm, flow rate was 0.4 mL/min with a total run time of 3 minutes and an injection volume of 6 μL. Recovery was >96% and calibration was linear between 0.25 and 30 μg/mL.

Pharmacokinetic Analysis

Pharmacokinetic parameters after a single oral dose of meloxicam (Part I) were determined for each horse by use of noncompartmental analysis with a commercial software program.10 Maximum concentration (Cmax) of meloxicam and time to Cmax (tmax) were read directly from the data. Terminal half-life (t½) was determined by log-linear regression. Area under the concentration time curve (AUC0∞∞) was calculated by log-trapezoidal integration. Drug accumulation ratio (R)[27] was calculated using the equation Rmax = CSSmax/C1max, where C1max was the peak concentration during the 1st dose interval (Part I) and CSSmax is the peak level at steady state (in this study, after the final dose at Day 41).

Statistical Analysis

Pharmacokinetic results from Part I and Part II are presented as descriptive data only. Age and body weight of horses in Treatment Groups A, B, C, D, and E in Part II of this study were compared by one-way analysis of variance. Gastroscopy squamous and glandular mucosa scores for each horse were compared before and at the completion of treatment by ordinal logistic regression, and the robustness of this analysis confirmed using logistic regression as previously described.[28] Clinical pathology results were evaluated for replicate and treatment group effects by separate two-way repeated measures analysis of variance, after checking for normal distribution and equal variance. Multiple pairwise comparisons were performed within and between groups when a significant time or treatment effect was determined using the Tukey′s test. Results are reported as mean ± SD (for descriptive data) or mean ± SEM (after ANOVA), and were considered significantly different for P < .05.



Maximum plasma concentrations (Cmax) after a single oral dose of meloxicam to fed horses ranged from 767.9 to 1113.2 ng/mL (mean 915.1 ± 116.9 ng/mL) (Fig 1). Tmax 2.62 ± 1.88 hours (range: 1.5–8 hours) and the area under the plasma concentration curve was 11 281 ± 3240 ng/hour/mL. The elimination half-life (t½) was 10.24 ± 3.04 hours.

Figure 1.

Plasma concentrations of meloxicam after a single oral dose (0.6 mg/kg). Results are given as mean (±SD) from 16 horses.

Plasma peak and trough concentrations of meloxicam, determined daily in Part II for Days 0–13 (Fig 2), demonstrated similar peak concentrations to those observed after a single oral dose of meloxicam in Part I. Prolonged administration was not associated with accumulation of the drug (Rmax = 1.11); Cmax and t½ determined after the final dose on Day 41 (1012.3 ± 309.7 ng/mL and 9.25 ± 2.64 hours, respectively) were similar to values obtained after a single dose (Part I).

Figure 2.

Peak and trough plasma meloxicam concentrations after repeated administration of meloxicam 0.6 mg/kg PO q24h for 14 days) and elimination profile after 6 weeks administration. Results are given as mean ± SD. Trough values were determined from samples collected immediately before the next dose; peak values were collected from samples collected 1.5 hours after administration, except on Days 0, 5, 13, and 41, when more frequent sampling was undertaken.

Plasma concentrations of meloxicam in horses receiving 1.8 mg/kg (Group C) and 3.0 mg/kg (Group D) were markedly higher than that was observed for horses receiving 0.6 mg/kg meloxicam daily PO (Fig 3). Peak plasma PBZ concentrations measured at Day 0 (16.64 ± 3.31 μg/mL) were similar to those reported previously,[29, 30] as were subsequent plasma PBZ concentrations (Fig 4).

Figure 3.

Plasma meloxicam concentrations after daily oral administration of 0.6 mg/kg (Group B), 1.8 mg/kg (Group C), or 3.0 mg/kg (Group D). Results are given as mean ± SD from samples collected immediately before treatment and 2, 4, 8, 12, and 24 hours after treatment on Days 0, 6, and 13.

Figure 4.

Plasma phenylbutazone concentrations after oral administration at 4.4 mg/kg PO q12h on Day 0, 2.2 mg/kg PO q12h for 4 days, and then 2.2 mg/kg PO q24h for 9 days. Results are given as mean ± SD from sample collected before treatment and 2, 4, 8, 12, and 24 hours after treatment on Days 0, 6, and 13.

Clinical Findings

A single oral dose of meloxicam at 0.6 mg/kg was well tolerated by adult horses in Part I of this study, with no horse demonstrating adverse effects. In Part II of the study, both Group A (control) and Group B horses (treated with 0.6 mg/kg meloxicam PO daily for 6 weeks) demonstrated no change in any of the clinical parameters assessed (physical examination, body weight, hematology, serum biochemistry, urinalysis, or bone marrow cytology) and no changes were appreciated on gastrointestinal or renal ultrasound. Fecal occult blood tests were positive for 4 Group A and 5 Group B horses on single or multiple occasions, but did not correlate with gastroscopic or ultrasound findings consistent with gastrointestinal ulceration.

One horse (M4) of 7 in Group C (meloxicam 1.8 mg/kg PO daily for 2 weeks) developed mild preputial edema associated with low plasma protein and albumin concentrations on Day 13 of the study. This condition resolved spontaneously on completion of treatment. Three other Group C horses (M20, M32, M33) demonstrated mild gastrointestinal signs including loose feces or reduced gastrointestinal sounds. Two (M20, M33) had poor appetites relative to other study horses. One of these horses (M33) was noted spending increased time recumbent by Day 13, and was euthanized for reasons unrelated to this trial after completion of the study. Postmortem examination demonstrated impaction of the right and left ventral colons. Gross and histological examination revealed mild ulceration of the squamous gastric mucosa, superficial erosions, and congestion of the glandular mucosa, consistent with gastroscopy findings in this horse. The right dorsal colon, kidneys, and liver demonstrated no significant changes.

Only two (M12, M30) of 7 Group D horses (meloxicam 3.0 mg/kg PO daily for 2 weeks) remained well for the duration of the study. One horse (M5) had decreased appetite within 24 hours of commencing treatment. By treatment Day 3, he was pyrexic and had profound leucopenia and neutropenia (3.1 × 109/L and 1.1 × 109/L, respectively; laboratory reference range for total leukocyte count 5.7 10.0 × 109/L and for neutrophils 2.8–8.0 × 109/L), with cells demonstrating normal morphology and no evidence of a left shift. This horse was withdrawn from the study at this time and bone marrow cytology, collected the following day, demonstrated a homogenous population of atypical large round cells containing large round nuclei with finely stippled chromatin and one or more prominent nucleoli. There were numerous mitotic figures associated with this population and these cellular changes were initially interpreted by clinical pathologists as suggestive of acute myeloid leukemia. Gastroscopic findings at this time demonstrated hyperemia of the glandular mucosa and a large discrete area of ulceration of the squamous mucosa at the lesser curvature. Ultrasound examination of the right dorsal colon was normal initially but thickening of the right dorsal colon was evident by Day 6. At this time (Day 6), the horse developed preputial and ventral edema associated with decreased serum total protein and albumin concentrations (55 and 24 g/L, respectively; reference ranges 55–75 g/L and 27–38 g/L). Total protein and albumin concentrations decreased to 45 and 18 g/L, respectively, 14 days after first treatment. The horse was managed with supportive care (high quality diet; oral or nasogastric administration of fluids and electrolytes) and made a full recovery over 8–10 weeks, without evidence of residual gastrointestinal dysfunction. Bone marrow cytology during convalescence demonstrated marked myeloid hyperplasia and a gradual return to normal cytology.

Two (of 7) other Group D horses (M7, M21) demonstrated soft or watery feces for a 24-hour period during the 1st week of the study. Both horses were recumbent for extended periods at these times, but demonstrated no other clinical signs. A 6th horse (M19) was observed to drink additional fluid toward the end of the study (Days 11–13), associated with watery feces. Ultrasound evaluation of this horse on Days 6 and 13 demonstrated a thickened right dorsal colon. The 7th Group D horse (M31) died acutely overnight on Day 13. Postmortem examination demonstrated peracute necrotizing typhlocolitis and acute renal medullary necrosis. Gross and histological inspection of other tissues, including liver and the remainder of the gastrointestinal tract revealed no further abnormalities.

Three of 5 Group E horses (receiving PBZ at the manufacturer's recommended dose rates) remained well for the duration of the study. One horse (M3) developed preputial edema associated with decreased serum total protein and albumin concentrations (Day 9), which resolved after completion of the study. The final Group E horse (M25) demonstrated a reduced appetite (from Day 2), was lethargic (from Day 4), intermittently pyrexic (rectal temperature 39.0°C), recumbent more often than other horses (from Day 5), and demonstrated an episode of moderate colic (pawing, rolling) which resolved without treatment on Day 6. Although her appetite remained decreased, she was otherwise comfortable until Day 13, when she again spent increased time in sternal or lateral recumbency and had cow pat feces. After completing the study (Day 16), this mare was euthanized for reasons unrelated to this trial and postmortem examination demonstrated marked submucosal thickening of the right dorsal colon involving a well circumscribed length of approximately 12 cm at the diaphragmatic flexure. Petecchial hemorrhages were evident on the serosal surface of the affected bowel, and there was diffuse ulceration of the mucosal surface. Pronounced edema of the submucosa and marked mixed inflammatory cell infiltration into the superficial submucosa and lamina propria were evident on histology. There was widespread ulceration and neutrophilic exudation of the mucosal surface. This lesion was located too far cranially to be visualized on ultrasound inspection of the right dorsal colon. Gross and histological inspection of the stomach, kidney, and liver demonstrated minor changes to the gastric mucosa only.

Abdominal Ultrasound

Weekly abdominal ultrasound examination demonstrated no abnormalities for Group A or Group B horses. Thickened right dorsal colon was evident in 2 of 7 Group D horses (M5, M19; meloxicam 3.0 mg/kg PO q24h) from Day 6. Although not evident in antemortem ultrasonographic examination, 1 (of 5) Group E horse (M25) also developed right dorsal colitis.


Results of gastroscopic examination are presented in Table 1. Logistic regression demonstrated no changes in squamous ulceration scores for any group. Within Group C, the odds of grade 2 ulceration of the glandular mucosa at the completion of the study were 4.72 times higher than for horses in Group A (control) (95% CI: 1.00, 22.33). In addition, only for horses in Group C, odds of glandular ulceration before treatment were approximately 83% lower than that was observed in the glandular mucosa after treatment (OR: 0.27, 95% CI: 0.098, 0.72).

Table 1. Results of gastroscopic examinations before treatment and after 2 weeks (Groups C, D, and E) or 6 weeks (Groups A and B) of treatment
 Squamous MucosaGlandular Mucosa
ScoreMedian ScoreTotal ObservationsScoreMedian ScoreTotal Observations
  1. Squamous mucosa scores were determined independently by 2 assessors blinded to treatment (0, intact epithelium with or without areas of reddening and hyperkeratosis; 1, small single or multifocal ulcers; 2, large single or multifocal lesions, or extensive superficial ulcers; 3, extensive, often coalescing ulcers with areas of deep ulceration). Glandular mucosa scores were determined by a single examiner blinded to treatment (Grade 0, intact mucosal epithelium, uniform, or relatively uniform color; Grade 1, obvious patchy erythema, multiple superficial erosions, or both; Grade 2, extensive superficial erosions (often circumferential or linear), with or without areas of deeper ulceration). Results are expressed as the frequency for each mucosal score, the median score, and the number of observations for each group before and after treatment.

  2. a

    One group D horse (M5) was withdrawn from treatment on Day 4.

Group A (placebo), n = 7
 Day 0931101451006
 Day 41833001451006
Group B (meloxicam 0.6 mg/kg q24h), n = 7
 Day 01013001451006
 Day 4162601143120.56
Group C (meloxicam 1.8 mg/kg q24h), n = 7
 Day 0860001431004
 Day 13482011421215
Group D (meloxicam 3.0 mg/kg q24h), n = 7a
 Day 0833001450005
 Day 13444011231105
Group E (PBZ 4.4 mg/kg q12h, 2.2 mg/kg q12h, 2.2 mg/kg q24h), n = 5
 Day 0460011040004
 Day 13262011013014

Clinical Pathology

Throughout Part II of the study, mean hematology and serum biochemistry results from Group A (control) horses and Group B (meloxicam 0.6 mg/kg PO q24h) horses did not change significantly. However, significant interactions were observed between treatment and time of collection for total leukocyte and neutrophil numbers (P < .001 and P = .002, respectively). Relative to Day 0 values, mean total white cell and neutrophil numbers were significantly reduced on Day 6 for both Groups C and D, but had returned to pretreatment levels by Day 16, 72 hours after completion of the study. Total white cell counts for Group E horses were significantly elevated on Day 13. Within each group, 3 (of 7) Group A (control) horses, 5 (of 7) Group C, all 7 Group D, and 2 (of 5) Group E horses demonstrated absolute leucopenia, neutropenia, or both on at least 1 occasion during the study. Red blood cell parameters, platelet numbers, and fibrinogen concentrations did not change significantly. Total protein, albumin, and globulin concentrations decreased significantly for Groups C, D, and E during the course of the study (Fig 5). Four (of 7) Group C, 6 (of 7) Group D, and 2 (of 5) Group E horses demonstrated total serum protein, albumin concentrations, or both below the laboratory reference range during the course of the study. Within each group, other serum biochemistry parameters did not change in a clinically significant way. There was no evidence of increased serum urea or creatinine in any individual horse nor in any treatment group, and no changes in liver enzyme activities or serum electrolyte concentrations. Urine enzyme concentrations appeared to vary randomly for horses in each group, and could not be related to time of collection or treatment. Protein concentration on dipstick evaluation of urine samples was consistently low, and abnormalities were not observed on examination of urine sediment.

Figure 5.

Total serum protein (a) and albumin (b) concentrations for all horses during the study. Results are expressed as mean ± SEM, calculated by two-way repeated measures analysis of variance with time and treatment as fixed effects. Significant interactions between time and treatment were observed (both P < .001). Results with the same superscript indicate a significant within group (time of collection) effect (P < .05). Significant differences between treatment and control groups are indicated (*P < .05; **P < .01).

Aspiration of bone marrow was initiated after the first Treatment Group D horse (M5) developed leucopenia and neutropenia. Hence, Day 0 samples were not available for 1 horse in each treatment group (horses M1–M5). Bone marrow cytology was unremarkable for all horses except M5 (Group D, meloxicam 3.0 mg/kg PO q24h). Myeloid:erythroid ratios were determined for all samples but have not been further analyzed as bone spicules were variably observed, suggesting that sample quality may have been inadequate in some instances.


Maximum plasma meloxicam concentrations observed in Parts I and II of the current study were less than those reported previously in fed or fasted horses,[31] although tmax and elimination half-life were similar. Plasma meloxicam concentrations in our studies after single and repeated doses of 0.6 mg/kg PO were between 200 and 1000 ng/mL, consistent with values reported to decrease lameness and joint inflammation,[15, 17] and to have anti-inflammatory effects in vitro.[18, 19] The same dose of meloxicam administered IV elicited postoperative analgesia comparable with that of flunixin.[16] As our calculated Rmax value of 1.1 was consistent with the predicted value for a drug being administered at a dosing interval approximately 2.5 times the elimination half-life,[27] there was no evidence of drug accumulation in our study.

Although the administration of repeated doses of meloxicam at the recommended dose rate for 6 weeks (Group B) was well tolerated by all horses, meloxicam administration at 3 or 5 times the recommended dose rate was associated with clinical findings consistent with gastrointestinal damage (10 of 14 horses), myeloid dyscrasia (1 horse), and renal damage (1 horse). One (of 7) Group C horse, receiving 3 times the recommended dose rate of meloxicam, had an impacted colon at completion of the study, but had not demonstrated clinical signs associated with colic (rolling, pawing, or elevated heart rate), suggesting that the drug might have masked discomfort associated with the accumulation of ingesta. Although it is possible that this horse had an impaction before commencing the current study and that the development of this condition was unrelated to drug treatment, it is also possible that intestinal motility was altered by the administration of meloxicam. Selective COX-2 inhibition has been associated with reduced tonic and spontaneous phasic small intestinal contractions.[32] Although nonselective COX inhibitors (flunixin and indomethacin) might have little or no effect on motility,[32] other studies have suggested that nonsteroidal agents could reduce colonic contractility.[33] Potential effects of meloxicam on intestinal motility have not been investigated in horses.

Clinical signs and clinical pathology changes were more severe in horses receiving 3.0 mg/kg of meloxicam (5 times the recommended dose), with 6 of 7 Group D horses demonstrating abnormalities on physical examination, clinical pathology aberrations, or both during the study. One horse in this group died peracutely with renal and gastrointestinal damage, and 2 other horses developed right dorsal colitis which resolved after discontinuing treatment. Human studies (reviewed by Warner and Mitchell[34]) have demonstrated that COX-2 selective drugs are associated with fewer gastrointestinal adverse effects than nonselective COX inhibitors, and inhibition of both COX-1 and COX-2 is required to induce acute gastrointestinal damage in animal models.[35] At high concentrations (greater than approximately 1200 ng/mL), meloxicam is also an effective inhibitor of equine COX-1,[19] hence, plasma concentrations achieved at higher doses in our study were likely to have inhibited both isoforms of the enzyme in Group C and D horses. Not surprisingly, adverse effects observed in these horses (gastric ulceration, right dorsal colitis, and protein losing enteropathy) were consistent with previous reports of NSAID toxicity in horses. Interestingly, despite numerous reports of myeloid dyscrasia in the medical literature associated with the administration of PBZ,[36-38] bone marrow changes have not been previously reported associated with NSAID administration in horses. Meloxicam has been reported as increasing myeloid cell production in mice because of increased production of GM-CSF.[39-41] Given the profound reduction in circulating white cell numbers, it is probable that changes in bone marrow cytology observed in horse M5 in the current study were associated with decreased release or maturation of undifferentiated cells. Reduced neutrophil numbers can also be associated with increased margination because of proinflammatory mediators or increased concentrations of endotoxin, which in this study might have been associated with NSAID induced changes in gastrointestinal mucosal integrity. However, comparable changes in bone marrow cytology were not observed in other horses with gastrointestinal changes, leucopenia, neutropenia, or both in the current study. Our impression was that peripheral leukocyte changes preceded clinical evidence of intestinal damage. Toxic changes and increased circulating numbers of immature neutrophils (band cells and metamyelocytes), typical of leukocyte changes induced by endotoxin or proinflammatory mediators, were not observed in peripheral blood leucogram of this horse until Day 9. Changes in peripheral blood and bone marrow neutrophils were reversed upon discontinuation of treatment, and resolved more quickly than did evidence of gastrointestinal mucosal damage. Myeloid changes or abnormal cell types were not recognized in any other study horse.

Leucopenia and neutropenia were the most common hematological findings with high doses of meloxicam. Serum biochemistry results demonstrated decreased total serum protein and albumin in 4 (of 7) Group C horses and 6 (of 7) Group D horses. Consistent with previous reports of NSAID toxicoses,[9, 42] this was a sensitive indicator of adverse drug effects. Renal protein loss did not appear to be a feature of drug toxicity in any group, and renal damage was documented in only 1 Group D horse (M33). Increased urine concentrations of GGT or ALP have been suggested as the most sensitive method to detect acute renal tubule damage,[43] but were not useful in the current study. This may suggest that renal damage did not occur, except terminally in 1 horse, or may relate to limitations in sample submission (overnight) to a commercial clinical pathology laboratory. Ultrasound evaluation has previously been useful for recognition of renal pathology in adult horses and foals associated with NSAID administration,[44, 45] but did not reveal evidence of renal damage in the current study. Urinalysis and sediment examination were also unremarkable. Apparent low concentrations of protein (<0.3 g/L), evident on reagent strip urinalysis, were not associated with abnormalities on sediment examination and likely represented false positive results in alkaline urine (urine pH consistently >7.5). Results of fecal occult blood tests did not correlate with gastroscopy or ultrasound examination findings, and were considered unreliable in this context.

Two of 5 horses receiving PBZ at the recommended dose rate[46-49] also developed hypoproteinemia because of significant decreases in albumin concentration, and right dorsal colitis was confirmed in 1 affected horse. This horse had a peak plasma PBZ concentration of 21.6 μg/mL on Day 0, higher than all other horses at all other sampling times. The age of this horse (20 years) might have contributed to her susceptibility to adverse drug effects. It has been suggested that aged horses eliminate PBZ more slowly than younger animals, therefore making them much more susceptible to the drug's toxic effect.[50] Although this mare did not demonstrate delayed elimination of the drug, it is possible that other age-related changes affected drug metabolism, excretion or effects of PBZ, or its active metabolite (oxyphenbutazone). Plasma concentrations of PBZ producing adverse effects remain unknown.[49]

Interindividual differences in pharmacokinetics, efficacy, or susceptibility to adverse effects of NSAIDs are well recognized, and might be attributable to polymorphism of genes controlling drug metabolism or response.[51] For example, differences in genes for the GPIIa platelet receptor might contribute to aspirin resistance in people[52] and variations in plasma concentrations of meloxicam in people have been associated with variation in the CYP2C9 gene.[53] Although not characterized in the horse, genes encoding functional differences in the cytochrome P450 2C subgroup likely contribute to NSAID elimination and it is possible that genetic differences mediate individual drug responses in this species, accounting for the relatively wide range of pharmacokinetic indices and the variable development of adverse effects within treatment groups in this study. However, human studies[54] (and reviewed by Rodrigues[55]) have failed to associate gastrointestinal adverse effects of NSAIDs to polymorphism of the CYP2C9 gene. Physiological factors might also affect drug absorption, metabolism, or elimination, and thereby contribute to variation in plasma concentrations and, potentially, efficacy, or adverse effects. For example, the bioavailability of meloxicam in horses is affected by feeding.[31] Although not reported in horses, studies in rats[56] have suggested that gender may influence the severity of adverse effects observed after treatment with aspirin or phenylbutazone, although strain differences were also observed.

There were no adverse effects associated with the repeated oral administration of meloxicam at 0.6 mg/kg for 6 weeks to healthy adult horses, although the administration of higher doses of meloxicam was associated with dose-dependent adverse effects typical of NSAIDs. The dose rate of 0.6 mg/kg every 24 hours achieved plasma concentrations likely to be therapeutic. PBZ at the recommended dose rate caused clinically significant right dorsal colitis in at least 1 treated horse (of 5) and significant protein loss (hypoproteinemia) in 2 horses.


Grant Support: This work was funded by Troy Laboratories Australia Pty Ltd. Dr Joe Pippia was employed by Troy Laboratories at the time study was conducted.

The assistance of Kellie Tinworth, Heather Clegg, Tara Campbell, Timna Dean, Fiona Schneiders, Sarah Hanlon, Kristie Hann, and Ken Jacobs is gratefully acknowledged. Equine science (Danielle Arthur, Naomi Bakker, Whitney Chapple, Freya Colvern, Fiona Edwards, Simone Healey, Amelia Pascoe, and Alecia Sheridan) and veterinary science students (Amanda-Lee Charman, Greg Dale, Alistair Grant, Shahid Khalfan, Trystan Keylock, Tara Mills, Emily Roberts, Anneliese Seagar, and Sarah Ward) assisted ably with sample collection and care of horses. HPLC was performed by Fiona Cotter. Gross and histopathology services were provided by Associate Professor Shane Raidal.


  1. 1

    Meloxicam Oral Suspension (30 mg/mL), Troy Ilium Pty Ltd, Smithfield, NSW

  2. 2

    Oralject P-Butazone Paste (200 mg/mL); Virbac Australia, Milperra, NSW

  3. 3

    Product vehicle only, supplied by Troy Ilium Pty Ltd

  4. 4

    Idexx Laboratories, Rydalmere, NSW

  5. 5

    Olympus CV160; Austvet Endoscopy, Melbourne, VIC

  6. 6

    MyLab70, Esote; distributed by Medical Plus, Crows Nest, NSW

  7. 7

    Hemoccult Sensa, Beckman Coulter Australia Pty Ltd, Gladesville, NSW

  8. 8

    Waters Corporation, Rydalmere, NSW

  9. 9

    Acquity BEH, Waters Corporation

  10. 10

    Topfit 2.0, Gustav Fischer Verlag