Vet therapeutics masterpage

M. Drag, B. N. Kunkle, D. Romano, and P. D. Hanson Efficacy of Firocoxib in Preventing Urate-Induced
Synovitis, Pain, and Inflammation in Dogs*
Marlene Drag, DVM, MS, DACLAM
Bruce N. Kunkle, DVM, MS, PhD
Davida Romano, MPH
Peter D. Hanson, DVM, PhD, DACVS
Merial Limited3239 Satellite BoulevardDuluth, GA 30096 This positive-control study evaluated the efficacy of firocoxib versus carprofen,deracoxib, and meloxicam for the prevention of pain and inflammation in a uratecrystal synovitis model of lameness. Lameness scoring and force plate gaitanalysis were used to assess efficacy. The resulting lameness scores and forceplate ground reaction forces after urate crystal injection were not significantlydifferent among the groups. Relative to each group’s baseline (nonlame) score,only the firocoxib group was not significantly lame, based on lameness score,at the model’s peak effect.
■ INTRODUCTION
tions, whereas COX-2 is thought to be respon- sible for inflammatory prostaglandin.3 This gesics for treating osteoarthritis (OA) in humans concept may be overly simplistic, but it has re- and animals.1,2 Veterinary NSAID use has blos- sulted in an intense search for agents that sup- somed over the past decade with the introduc- tion of products featuring improved safety pro- products became available, studies were con- files compared with aspirin, the primary NSAID ducted to compare NSAIDs and their ability to used in dogs before the mid-1990s. Gastroin- inhibit COX in vitro. This ability is frequently testinal side effects often limited the duration of expressed as the concentration ratio (COX-1: aspirin treatment, frustrating those treating os- COX-2) that results in the inhibition of 50% teoarthritic dogs. Research to improve NSAID safety and efficacy has been fervent and often fo- spectively).4–7 These studies provided much- cused on investigating cyclooxygenase (COX) needed information regarding techniques for activity of NSAIDs with therapeutic potential.
preliminary evaluation of new NSAIDs, but COX-1 is the isoform thought to be respon- the results should not be overinterpreted, espe- sible for basal physiologic prostaglandin func- cially when extrapolating to clinical situations.
*This research was funded by Merial Limited, Du- Differences in COX-1–sparing ratios in vitro do not necessarily correlate with differences in Veterinary Therapeutics • Vol. 8, No. 1, Spring 2007 Practitioners can choose from many NSAIDs TABLE 1. Treatment Dosage
in several classes to provide safe and effective control of the pain associated with OA.1 This study evaluated the relative efficacy of a new veterinary NSAID, firocoxib (Previcox, Merial [ML-1,785,713]). Of the drugs compared inthis study, firocoxib and deracoxib (Deramaxx, Novartis) belong to the coxib class of nonnar- cotic NSAIDs, carprofen (Rimadyl, Pfizer Ani- mal Health) belongs to the propionic acid class, aDosages were based on body weight and rounded to and meloxicam (Metacam, Boehringer Ingel- the nearest half-tablet increment to ensure that at least heim Vetmedica) belongs to the oxicam class.
the labeled dosage was achieved for carprofen, dera- These drugs are indicated for either the relief or coxib, and firocoxib. For meloxicam, the liquid sus- control of pain and inflammation associated pension was dosed to the exact body weight.
with OA in dogs. The NSAIDs selected forcomparison with firocoxib were chosen because prostaglandin production in vivo,8 but the in- they have achieved high levels of acceptance by ability of nonselective NSAIDs to spare COX- 1 has been shown to be a major factor con- This study evaluated the efficacy of these tributing to adverse effects associated with NSAIDs to prevent lameness associated with a their use.9 Obviously, the decisive NSAID se- UC lameness model. This method of acute syn- lection points, as with any drug, are safety and ovitis induction in dogs and humans was first documented in 1962.11 In that study, the au- Firocoxib is a highly selective inhibitor of thors injected themselves with 20 mg of UC in COX-2 developed specifically for veterinary the knee and followed the progression of in- use. Its efficacy and safety were demonstrated flammation and pain. The injection of UCs elic- Firocoxib is a highly selective inhibitor of COX-2
developed specifically for veterinary use.
in support of its approved indication for con- its an intense inflammatory response, which in- trol of pain and inflammation associated with cludes neutrophil migration and release of such inflammatory mediators as prostaglandins, study found that firocoxib has 380-fold selec- leukotrienes, and others. The model has been tivity for COX-2 over COX-1 (more precisely, refined over the years and has been widely used the IC ratio of COX-1:COX-2 was calculat- to assess NSAID efficacy.10–18 Although the in- ed as 384) and, with the use of the urate crys- flammatory response is an acute process, many tal (UC) model of lameness, demonstrated that components are similar to those that occur in firocoxib provided greater efficacy than carpro- chronic conditions. In dogs, lameness is evident fen whether given as a preventive or as a treat- within 2 hours after UC injection, reaches a peak at approximately 4 hours, and then gradu- M. Drag, B. N. Kunkle, D. Romano, and P. D. Hanson TABLE 2. Treatment Schedulea
dogs return to full weight-bearingcapacity by 24 to 36 hours after injection makes it a suitable mod-el to evaluate NSAID efficacy. aDogs were assigned to replicates based on body weight and randomly as- signed to sequences within replicates.
1–4 = treatment group; L = left stifle; R = right stifle.
cam) in the UC lameness model.
Comparing the firocoxib groupwith three NSAID positive-control groups pro- model is representative of the more intense end vided a humane method of evaluating firocoxib of the pain and inflammatory spectrum possi- efficacy without the need for an untreated neg- For allocation, the four heaviest dogs were as- signed to replicate 1 and the remaining four ■ MATERIALS AND METHODS
dogs were assigned to replicate 2. Within each Study Animals and Treatment Allocation
replicate, each dog was randomly allocated to Eight healthy mixed-breed dogs (four males one of four treatment sequences. By the end of and four females), approximately 70 months of the study, each treatment had been assessed in age and weighing 15 to 28 kg, were used in this each of the eight dogs. The UC injections were positive-control, blinded, four-period cross- alternated between left and right stifles for each over study using a randomized block design treatment period. Treatment periods were sepa- based on pretreatment body weights. All ani- rated by a 7-day washout period. The details of mals were managed similarly, with due regard the treatment schedule are outlined in Table 2.
for their well-being, and were handled in com-pliance with Merial Institutional Animal Care Lameness Model and Assessment
and Use Committee approvals. Carprofen, de- racoxib, firocoxib, and meloxicam were admin- traarticular injection of monosodium UC has istered orally at the approved label dosages for been described elsewhere.10,17,18 Briefly, the UC control of pain and inflammation associated suspension was prepared fresh for each dog by with OA (Table 1).19–22 The dosage of products available as a tablet formulation (i.e., carpro- fen, deracoxib, and firocoxib) was rounded to cating for 20 minutes. Dogs were anesthetized the nearest half-tablet increment that provided with propofol intravenously before intraarticu- at least the labeled dosage. For deracoxib, the lar injection of UC. The skin over the stifle dosage for OA differs from that for acute sur- joint was clipped and prepared for an aseptic gical pain; the OA dose was used because the procedure. A 20-gauge, 1.5-inch needle was in- other drugs were administered based on their serted into the joint. Once synovial fluid was evident, 1 ml of suspension containing 19 mg Veterinary Therapeutics • Vol. 8, No. 1, Spring 2007 TABLE 3. Procedure Schedule
monosodium UC was injected into the joint.
Lameness score and force plate gait analysis of The stifle was then extended and flexed several times to distribute the UC suspension.
and 8 hours after UC injection, respectively).
Baseline lameness score and force plate gait Following the hour 18 assessment, rescue med- analysis of ground reaction force (GRF) were ication was allowed for any dog that the inves- performed the day before treatment within 24 tigator judged to need it. The schedule of these hours of lameness induction. NSAID treat- ment was given at hour 0, the start of each Force plate gait analysis of GRF was per- treatment period. Lameness was induced by formed using a floor-mounted system to deter- UC injection of the stifle 10 hours after mine the peak vertical force (PVF) and vertical NSAID administration to assess treatment effi- impulse (VI). The equipment (Model OR6-7- cacy during the middle to end of the dosing 1000 Biomechanics Force Platform, Advanced Mechanical Technology, Watertown, MA) andmethodology were similar to those used in oth- 0 = No lameness
er studies analyzing GRF to evaluate and quan- 1 = Mild lameness, including toe touch to
tify the degree of lameness.17,23,24 The target or floor (i.e., weight bearing) on all of the goal was to make at least six valid observations for the affected rear leg at each time point 2 = Moderate lameness, including toe touch
for every dog. Dogs were trotted at a targeted velocity range of 1.5 to 3.0 m/sec and an acceleration of 0 ± 1.0 m/sec2. If the dog had 3 = Severe lameness, including toe touch to
non–weight-bearing lameness, force plate gait analysis was not performed and a value of 0.0 4 = Non–weight-bearing lameness, including
was recorded. The primary endpoint of the toe touch to floor on less than 50% of the force plate gait analysis was determination of the weight-bearing PVF of the affected limb, M. Drag, B. N. Kunkle, D. Romano, and P. D. Hanson TABLE 4. Individual Mean Velocities during Force Plate Gait Analysis
Mean Velocity
Velocity (m/sec)
(m/sec) across
Dog ID
Periods
Range
among the treatments were tested comparing weight (N/kg) and expressed as a percentage of firocoxib with each of the other three treat- the baseline PVF. A secondary endpoint was ments. Each treatment baseline PVF value was determination of VI, which was measured in also compared with the same treatment’s 4-hour Newton-seconds/kg body weight (N-sec/kg). post-UC injection value and expressed as a per- All personnel making efficacy observations centage of baseline. Each comparison was tested were blinded to treatment. The person ad- at a two-sided significance level of P < .05. ministering the treatments was not blindedbut did not participate in making any efficacy ■ RESULTS
Baseline lameness score was 0 for each group and time period. Baseline force plate gait analy- Statistical Analysis
sis values for each dog and leg were not signifi- A separate analysis was performed at each cantly different, with an average decrease in time point (baseline, hour 14, and hour 18) us- PVF of 2.2% from the first to second injection ing analysis of variance for a crossover design in the same leg. Although a relatively broad on a model including treatment, sequence, pe- range of velocities was permitted to accommo- riod, and carryover effects as fixed effects and date the individual dogs, the overall mean ve- replicate and the interaction of replicate and locity was 2.13 m/sec and the mean difference treatment as random effects. Preliminary statis- in velocities for a given dog across the four pe- tical analyses revealed there was no carryover riods was 0.21 m/sec (range: 0.12–0.40 m/sec; effect for any variables in the lameness scores Table 4). The mean acceleration was 0.025 or the force plate gait analysis. As such, a re- duced statistical model including treatment, sequence, and period as fixed effects and repli- during the peak of the UC-induced lameness cate and the interaction of replicate and treat- effect (4 hours postinjection), there was a sig- ment as random effects was used. Differences nificant difference between the lameness score Veterinary Therapeutics • Vol. 8, No. 1, Spring 2007 TABLE 5. Mean (±SD) Acceleration Values (m/sec)
by Study Period
–0.005 ± 0.456 –0.004 ± 0.508 0.020 ± 0.424 analysis of GRF and the result-ant percentage of baseline PVF values, the firocoxib group had the highest nu- TABLE 6. Lameness Score Results—
meric scores (most weight-bearing capacity) Treatment vs. Baseline
for least squares mean. The percentage of base- Hour 14
line PVF in the firocoxib group was 84.5 at (4 hr after UC injection)
hour 14 compared with a range of 53.2 to 63.2for the other groups (P = .078 to .170, Figure Treatment
2). Similarly, at hour 18 the percentage of base- line PVF in the firocoxib group was 94.5 com- pared with a range of 75.6 to 87.2 for the oth- er groups (P = .179 to .536).
Regarding force plate gait analysis of GRF and the resultant percentage of baseline VI val- ues, at hour 14 the firocoxib group value was P values are the contrast between the treatment group 76.1 compared with a range of 48.7 to 59.0 forthe other groups (P = .094 to .228). Similarly, athour 18 the firocoxib group percentage of base- reported for each of the carprofen, deracoxib, line VI was 89.3 compared with a range of 77.9 and meloxicam groups relative to that treat- to 87.4 for the other groups (P = .421 to .887).
ment group’s baseline (nonlame) score, but thedifference in the lameness scores between the ■ DISCUSSION
firocoxib-treated group and its baseline was vided relief throughout the day, the lameness challenge was initiated 10 hours after a single lameness and more weight-bearing capacity in dose of NSAID was administered and effects the firocoxib-treated group, there were no sig- were assessed at 14 and 18 hours after treat- ment. Considering the rigorous challenge pro- groups for either lameness score or force plate vided by the UC lameness model (intraarticu- gait analysis based on a mixed model analysis lar injection of 19 mg UC), all of the NSAIDs of variance for a four-period crossover design.
provided a positive benefit compared with his- Regarding lameness score of the firocoxib group versus the comparative NSAIDs, at hour In an effort to evaluate firocoxib using the 14 the firocoxib group had the lowest lameness most humane methods possible, there was no score, 0.75, compared with a range of 1.63 to negative-control group, a situation considered 2.0 for the meloxicam, deracoxib, and carpro- and evaluated here in depth as a potential fen groups (P = .057 to .125; Figure 1). This study deficiency. In lieu of a control group, one trend was maintained at hour 18, with the firo- can look to other studies to see what has typi- M. Drag, B. N. Kunkle, D. Romano, and P. D. Hanson Lameness Scores Were Lowest in the Firocoxib Group
Firocoxib (5 mg/kg)Deracoxib (1–2 mg/kg)Carprofen (4.4 mg/kg) Time after Treatment
Figure 1. Lameness score results (mean ± SEM; range: 0 [normal] to 4 [non–weight-bearing lameness]). Compar-
ison between firocoxib and comparative NSAIDs had P values of .057 (carprofen and deracoxib) to .125 (meloxi-
cam) at hour 14 and .181 (deracoxib) to .450 (carprofen) at hour 18.
cally been reported when lameness is induced the weak challenge of control subjects and sub- by UC injection in dogs. The same lameness sequent insufficient differentiation between model, using 19 mg sodium urate, was used in control and treatment. An estimate of the per- a ketoprofen dose study in which the placebo centage of baseline PVF scores in a study using control group had percentage of baseline PVF scores of 19.0 and 44.4 at 2 and 6 hours after weak challenge.16 Values taken from a chart of lameness induction, respectively.13 The per- PVF as a percentage of body weight and recal- centage of baseline PVF scores in the placebo culated as percentage of baseline PVF reveal control group in another study using the same that the placebo group had a value of approxi- model but 17 mg of sodium urate was 0.0 at mately 56% of baseline PVF at 2 hours after both 4 and 8 hours after UC injection, and the UC injection; the figure climbed to about 71% duration of the lameness effect lasted as long as at 4 hours and about 81% at 8 hours. In con- trolled studies with limited sample size, it is hours.17 In another NSAID efficacy study, the difficult to demonstrate a meaningful effect percentage of baseline PVF scores for the when the PVF in the untreated control group placebo control group was 0.0 at 4 hours after after lameness induction is close to the baseline injection of 15 mg sodium urate, 10.3 at 6 PVF and that of treatment groups at the time Using even lower doses of UC has been as- sociated with less-consistent results because of lameness challenge model that is sufficient to Veterinary Therapeutics • Vol. 8, No. 1, Spring 2007 Weight-Bearing Capacity Was Highest in the Firocoxib Group
Firocoxib (5 mg/kg)Deracoxib (1–2 mg/kg)Carprofen (4.4 mg/kg)Meloxicam (0.2 mg/kg) F
V
P
e
lin
e
ta
n
e
rc
e
P
Time after Treatment
Figure 2. Percentage of baseline peak vertical force (PVF) (mean ± SEM). Comparisons between firocoxib and com-
parative NSAIDs had P values of .078 (carprofen) to .170 (meloxicam) at hour 14 and .179 (deracoxib) to .536
(carprofen) at hour 18.
demonstrate therapeutic effect. This was illus- trated during preclinical and dose titration As previously mentioned, although all the prophylactic efficacy studies with firocoxib NSAIDs tested in the current study appeared to that included vehicle control groups. The re- provide benefit against this lameness challenge, sultant scores for percentage of baseline PVF in only the firocoxib group had no significant dif- the untreated control group was 0.0 at 4 hours ference between the lameness scores at baseline after UC injection and ranged from 0 to 6.9 at (nonlame) and at hour 14 (4 hours after UC in- 8 hours.10,25 In contrast, values in the firocoxib jection). Similarly, in multicenter, double-blind, group receiving the therapeutic dose of 5 randomized clinical field studies in dogs with mg/kg was greater than 70% of baseline PVF OA, firocoxib demonstrated greater levels of im- at both time points. Evaluation of negative- provement in lameness compared with etodolac control groups in these studies provided ample and carprofen when the NSAIDs were adminis- evidence of the reproducible nature of the chal- tered for 30 days.26,27 These results suggest that it lenge caused by the reversible UC lameness is possible to observe differences in the efficacy re- model using 19 mg of UC for intraarticular in- sponse with different NSAIDs and that the UC jection and alleviated the need for a negative- lameness model is a predictor of such difference.
M. Drag, B. N. Kunkle, D. Romano, and P. D. Hanson ■ CONCLUSION
nary medicine. J Vet Intern Med 19(5):633–643, 2005.
These data demonstrate that the ability of 10. McCann ME, Andersen DR, Zhang D, et al: In vitro effects and in vivo efficacy of a novel cyclooxygenase- firocoxib to prevent pain in a UC lameness 2 inhibitor in dogs with experimentally induced syn- model compares favorably with that of carpro- ovitis. Am J Vet Res 65(4):503–512, 2004.
fen, deracoxib, and meloxicam. Although there 11. Faires JS, McCarty DJ: Acute arthritis in man and dog was no significant difference in efficacy among after synovial injection of sodium urate crystals.
groups, only the firocoxib group had no signif- icant difference between lameness score at the 12. Hazewinkel HA, van den Brom WE, Theijse LF, et al: Reduced dosage of ketoprofen for the short-term and peak of the lameness effect and the group’s long-term treatment of joint pain in dogs. Vet Rec 13. Bonneau S, Najbar W, Sanquer A, et al: Analgesic ef- ■ ACKNOWLEDGMENT
ficacy of nimesulide in a canine osteoarthritis model.
Rev Med Vet 156(4):179–181, 2005.
The authors thank Tad B. Coles, DVM, of Overland Park, 14. Borer LR, Peel JE, Seewald W, et al: Effect of carpro- Kansas, for technical assistance with this manuscript.
fen, etodolac, meloxicam, or butorphanol in dogswith induced acute synovitis. Am J Vet Res 64(11): ■ REFERENCES
1. Lascelles BD, McFarland JM, Swann H: Guidelines 15. Millis DL, Weigel JP, Moyers T, Buonomo FC: Effect for safe and effective use of NSAIDs in dogs. Vet Ther of deracoxib, a new COX-2 inhibitor, on the preven- tion of lameness induced by chemical synovitis in 2. Johnston SA, Budsberg SC: Nonsteroidal anti-inflam- dogs. Vet Ther 3(4):453–464, 2002.
matory drugs and corticosteroids for the management 16. Cross AR, Budsberg SC, Keefe TJ: Kinetic gait analy- of canine osteoarthritis. Vet Clin North Am Small sis assessment of meloxicam efficacy in a sodium Anim Pract 27(4):841–862, 1997.
urate-induced synovitis model in dogs. Am J Vet Res 3. Jones CJ, Budsberg SC: Physiologic characteristics and clinical importance of the cyclooxygenase isoforms in 17. Rumph PF, Kincaid SA, Baird DK, et al: Vertical dogs and cats. JAVMA 217(5):721–729, 2000.
ground reaction force distribution during experimen- 4. Wilson JE, Chandrasekharan NV, Westover KD, et al: tally induced acute synovitis in dogs. Am J Vet Res Determination of expression of cyclooxygenase-1 and -2 isozymes in canine tissues and their differential sen- 18. Schumacher HR, Phelps P, Agudelo CA: Urate crystal sitivity to nonsteroidal anti-inflammatory drugs. Am J induced inflammation in dog joints: Sequence of syn- Vet Res 65(6):810–818, 2004.
ovial changes. J Rheumatol 1(1):102–113, 1974.
5. Kay-Mugford P, Benn SJ, LaMarre J, Conlon P: In 19. Food and Drug Administration: Freedom of Information vitro effects of nonsteroidal anti-inflammatory drugs Summary S/NADA 141-111 Rimadyl® (carprofen) on cyclooxygenase activity in dogs. Am J Vet Res Chewable Tablets for Dogs. November 26, 2001; ac- cessed July 2006 at www.fda.gov/cvm/FOI/foiabst2.html.
6. Brideau C, Van Staden C, Chan CC: In vitro effects 20. Food and Drug Administration: Freedom of Informa- of cyclooxygenase inhibitors in whole blood of horses, dogs, and cats. Am J Vet Res 62(11):1755–1760, 2001.
able Tablets (firocoxib). July 21, 2004; accessed July 7. Streppa HK, Jones CJ, Budsberg SC: Cyclooxygenase 2006 at www.fda.gov/cvm/FOI/foiabst2.html.
selectivity of nonsteroidal anti-inflammatory drugs in 21. Food and Drug Administration: Freedom of Informa- canine blood. Am J Vet Res 63(1):91–94, 2002.
tion Summary NADA 141-213 Metacam® (meloxi- 8. Sessions JK, Reynolds LR, Budsberg SC: In vivo ef- cam) 0.5 mg/ml and 1.5 mg/ml Oral Suspension. fects of carprofen, deracoxib, and etodolac on April 15, 2003; accessed July 2006 at www.fda.gov/ prostanoid production in blood, gastric mucosa, and synovial fluid in dogs with chronic osteoarthritis. Am 22. Food and Drug Administration: Freedom of Informa- J Vet Res 66(5):812-817, 2005.
tion Summary Supplemental NADA 141-203 DE- 9. Bergh MS, Budsberg SC: The coxib NSAIDs: Poten- RAMAXXTM Chewable Tablets (deracoxib). February tial clinical and pharmacologic importance in veteri- 11, 2003; accessed July 2006 at www.fda.gov/cvm/ Veterinary Therapeutics • Vol. 8, No. 1, Spring 2007 23. Rumph PF, Kincaid SA, Visco DM, et al: Redistribu- 26. Hanson PD, Brooks KC, Case J, et al: Efficacy and tion of vertical ground reaction force in dogs with ex- safety of firocoxib in the management of canine os- perimentally induced chronic hindlimb lameness. Vet teoarthritis under field conditions. Vet Ther 7(2):127– 24. Budsberg SC: Long-term temporal evaluation of 27. Pollmeier M, Toulemonde C, Fleishman C, Hanson ground reaction forces during development of experi- PD. Clinical evaluation of firocoxib and carprofen for mentally induced osteoarthritis in dogs. Am J Vet Res the treatment of dogs with osteoarthritis. Vet Rec

Source: http://www.traumatologiaveterinaria.com/articuls/docs/VTX-Spring07-Hanson.pdf