<%BANNER%>

Pharmacokinetic and Duration of Cyclooxygenase Inhibition Studies of Phenylbutazone, Ketoprofen and Flunixin Meglumine i...

Permanent Link: http://ufdc.ufl.edu/UFE0043833/00001

Material Information

Title: Pharmacokinetic and Duration of Cyclooxygenase Inhibition Studies of Phenylbutazone, Ketoprofen and Flunixin Meglumine in Athletic Thoroughbred Horses
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Hatzel, Jennifer N
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: doping -- horseracing -- nsaids -- pharmacokinetics
Veterinary Medicine -- Dissertations, Academic -- UF
Genre: Veterinary Medical Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Non-steroidal anti-inflammatory drugs (NSAIDs) represent a family of therapeutic agents that are one of the most widely administered medications in both human and veterinary medicine. Equine veterinarians have been particularly keen of utilizing their anti-inflammatory actions on anything from colic, to endotoxemia or to alleviate pain associated with a musculoskeletal injury. The Thoroughbred racing industry has historically represented equine athletes at the peak of their performance abilities and consequently at a great risk for injuries associated with this type of athleticism. A variety of NSAIDs such as flunixin meglumine, phenylbutazone and ketoprofen, have historically been administered to Thoroughbred racehorses in an effort to alleviate pain and reduce inflammation, ultimately with anticipation for continued racing. While NSAID administration for relief of pain is humane and necessary, racing authorities around the world prohibit the unscrupulous use of these drugs to enable a compromised horse to race. The first objective described in these experiments was to characterize the pharmacokinetics of flunixin meglumine, phenylbutazone and ketoprofen in conditioned Thoroughbred horses utilizing state of the art LC/MS/MS and determine updated pharmacokinetic parameters for each drug. The second objective described in these experiments aimed to determine the extent and duration of cyclooxygenase inhibition, after a single dose of each drug to evaluate the inhibitory capabilities for an active inflammatory event. Finally, the data obtained from these experiments was utilized in creating a novel method of determining an estimated time of administration and potentially enhance our knowledge of withdrawal times for each of these drugs. In order to obtain this data 6 Thoroughbred horses (3 mares and 3 geldings) in athletic condition sufficient to work 1 mile in 2 minutes without undue stresses were included. Each horse in the study was administered a single intravenous dose of one NSAID at the beginning of the week, while remaining in training. Blood samples were collected one hour and immediately prior to drug administration. Following administration of the NSAID, collections took place at 15 and 30 minutes, 1, 2, 3, 4, 6, 8, 24, 48 and 72 hours. Plasma samples were obtained following centrifugation and placed in -80°C storage until further processing by the Florida Racing Laboratory for pharmacokinetic analysis. Plasma and serum was obtained following an appropriate incubation period for each sample and stored at -80°C until further processing. All samples were subjected to PGE2 concentration (plasma) or TXB2 concentration (serum) analysis through the use of a commercially available ELISA kit. The pharmacokinetic data obtained for each NSAID provided an updated overview, along with the ability to detect increasingly minute concentrations of drug, not previously determined. Challenged by noxious stimuli, phenylbutazone was significantly inhibitory (p<0.5) towards TXB2 for 48 and PGE2 for up to 72 hours; ketoprofen demonstrated inhibitory actions on both mediators for up to 48 hours; and flunixin meglumine significantly inhibited both inflammatory mediators up to 8 hours. By combining the data obtained from these studies, a statistical model was created in an effort to use both the concentration of drug in a plasma sample along with the inhibited concentrations of either inflammatory mediatory (PGE2 or TXB2) to determine an estimated time of administration. The increasing sensitivity of analytical techniques along with the inability to completely correlate a plasma concentration with clinical efficacy, suggests a need to create novel analytical methods to aid in governing the anti-doping industry in Thoroughbred racing. The current study suggests a statistical model utilizing both pharmacokinetic and pharmacodynamic parameters to estimate an administration time. Potentially, this method could be used to enhance our existing knowledge regarding withdrawal times and in turn, better educate veterinarians and trainers towards the consequences of administering a monitored NSAID within a more scientifically established window of time prior to post.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jennifer N Hatzel.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Colahan, Patrick T.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043833:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043833/00001

Material Information

Title: Pharmacokinetic and Duration of Cyclooxygenase Inhibition Studies of Phenylbutazone, Ketoprofen and Flunixin Meglumine in Athletic Thoroughbred Horses
Physical Description: 1 online resource (116 p.)
Language: english
Creator: Hatzel, Jennifer N
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: doping -- horseracing -- nsaids -- pharmacokinetics
Veterinary Medicine -- Dissertations, Academic -- UF
Genre: Veterinary Medical Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Non-steroidal anti-inflammatory drugs (NSAIDs) represent a family of therapeutic agents that are one of the most widely administered medications in both human and veterinary medicine. Equine veterinarians have been particularly keen of utilizing their anti-inflammatory actions on anything from colic, to endotoxemia or to alleviate pain associated with a musculoskeletal injury. The Thoroughbred racing industry has historically represented equine athletes at the peak of their performance abilities and consequently at a great risk for injuries associated with this type of athleticism. A variety of NSAIDs such as flunixin meglumine, phenylbutazone and ketoprofen, have historically been administered to Thoroughbred racehorses in an effort to alleviate pain and reduce inflammation, ultimately with anticipation for continued racing. While NSAID administration for relief of pain is humane and necessary, racing authorities around the world prohibit the unscrupulous use of these drugs to enable a compromised horse to race. The first objective described in these experiments was to characterize the pharmacokinetics of flunixin meglumine, phenylbutazone and ketoprofen in conditioned Thoroughbred horses utilizing state of the art LC/MS/MS and determine updated pharmacokinetic parameters for each drug. The second objective described in these experiments aimed to determine the extent and duration of cyclooxygenase inhibition, after a single dose of each drug to evaluate the inhibitory capabilities for an active inflammatory event. Finally, the data obtained from these experiments was utilized in creating a novel method of determining an estimated time of administration and potentially enhance our knowledge of withdrawal times for each of these drugs. In order to obtain this data 6 Thoroughbred horses (3 mares and 3 geldings) in athletic condition sufficient to work 1 mile in 2 minutes without undue stresses were included. Each horse in the study was administered a single intravenous dose of one NSAID at the beginning of the week, while remaining in training. Blood samples were collected one hour and immediately prior to drug administration. Following administration of the NSAID, collections took place at 15 and 30 minutes, 1, 2, 3, 4, 6, 8, 24, 48 and 72 hours. Plasma samples were obtained following centrifugation and placed in -80°C storage until further processing by the Florida Racing Laboratory for pharmacokinetic analysis. Plasma and serum was obtained following an appropriate incubation period for each sample and stored at -80°C until further processing. All samples were subjected to PGE2 concentration (plasma) or TXB2 concentration (serum) analysis through the use of a commercially available ELISA kit. The pharmacokinetic data obtained for each NSAID provided an updated overview, along with the ability to detect increasingly minute concentrations of drug, not previously determined. Challenged by noxious stimuli, phenylbutazone was significantly inhibitory (p<0.5) towards TXB2 for 48 and PGE2 for up to 72 hours; ketoprofen demonstrated inhibitory actions on both mediators for up to 48 hours; and flunixin meglumine significantly inhibited both inflammatory mediators up to 8 hours. By combining the data obtained from these studies, a statistical model was created in an effort to use both the concentration of drug in a plasma sample along with the inhibited concentrations of either inflammatory mediatory (PGE2 or TXB2) to determine an estimated time of administration. The increasing sensitivity of analytical techniques along with the inability to completely correlate a plasma concentration with clinical efficacy, suggests a need to create novel analytical methods to aid in governing the anti-doping industry in Thoroughbred racing. The current study suggests a statistical model utilizing both pharmacokinetic and pharmacodynamic parameters to estimate an administration time. Potentially, this method could be used to enhance our existing knowledge regarding withdrawal times and in turn, better educate veterinarians and trainers towards the consequences of administering a monitored NSAID within a more scientifically established window of time prior to post.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jennifer N Hatzel.
Thesis: Thesis (M.S.)--University of Florida, 2011.
Local: Adviser: Colahan, Patrick T.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043833:00001


This item has the following downloads:


Full Text

PAGE 1

1 PHARMACOKINETIC AND DURATION OF CYCLOOXYGENASE INHIBITION STUDIES OF PHENYLBUTAZONE, KETOPRO FEN AND FLUNIXIN MEGLUMINE IN ATHLETIC THOROUGHBRED HORSES By JENNIFER NOELLE HATZEL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Jennifer Noelle Hatzel

PAGE 3

3 To the horses, without which, non e of this would be possible

PAGE 4

4 ACKNOWLEDGMENTS I thank my committee member s: Dr. Patrick Colahan, for the initial concept and subsequent support through the Racing Medi cation Testing Consortium (RMTC) for completion of these studies, Dr. Alison Mo rton for her constant encouragement and Dr. Tom Vickroy for his thoughts and ideas, always in the nick of time. I thank Dr. David Hurley and Natalie Norton from the University of Georgia for thei r patient guidance in assisting me with learning to perform the assays. I thank Dr. Richard Sams for bestowing me with a very basic knowledge of pharmacokinetics and Marc Rumpler for his unwavering support and mentorship through ev ery aspect of this process. I thank Dan Neal, from the department of biostatist ics for not only teaching me the basics, but also assisting me when I was way over my head. This project would have never been possible without the help of the entire staff of the Univ ersity of Florida Equine Performance Laboratory: Brett Rice, Allie Hreha, Amber Davidson and Megan Davidson, for whom I am very grateful. Finally, I owe a great deal of support and encouragement from Jeremiah, always.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 8 LIST OF FI GURES .......................................................................................................... 9 LIST OF EQUATIONS ................................................................................................... 10 LIST OF ABBREVI ATIONS ........................................................................................... 11 ABSTRACT ................................................................................................................... 12 CHAPTER 1 INTRODUC TION .................................................................................................... 15 The Eicosa noids ..................................................................................................... 15 History .............................................................................................................. 15 The Inflammatory Cascade .............................................................................. 16 Pharmaceutical Intervent ion ................................................................................... 19 The History and Mechani sm of N SAIDs ........................................................... 19 Adverse Effects Affiliated wit h NSAID Admini stration ....................................... 22 Novel Classes of Se lective N SAIDs ................................................................. 23 NSAID Use in Veteri nary Medi cine ......................................................................... 26 Equine NSAID Administration ........................................................................... 26 Drug Doping in the Pe rformance Horse ........................................................... 29 Monitoring Dr ug Abus e ..................................................................................... 31 Summary and Ob jectives ........................................................................................ 37 2 THE PHARMACOKINETICS OF PH ENYLBUTAZONE, KETOPROFEN AND FLUNIXIN MEGLUMINE FO LLOWING A SINGLE AD MINISTRATION ................. 41 Background ............................................................................................................. 41 Materials and Methods ............................................................................................ 42 Animals ............................................................................................................. 42 Standard Training Regimen for the Un iversity of Florida Equine Performanc e Lab ........................................................................................... 42 Incremental Exercise Te st to Exhaus tion ......................................................... 43 Standard Condition Test to Verify t he Ability to Gallop One Mile in Two Minutes .......................................................................................................... 43 Drug Administration and Sa mple Collec tion ..................................................... 44 Drug Analys is .......................................................................................................... 44 Chemicals and R eagents ................................................................................. 44 Phenylbutazone ................................................................................................ 45

PAGE 6

6 Sample pre paration .................................................................................... 45 Instrument ation .......................................................................................... 46 Data analys is ............................................................................................. 47 Ketoprof en ........................................................................................................ 48 Sample pre paration .................................................................................... 48 Instrument ation .......................................................................................... 49 Data analys is ............................................................................................. 50 Flunixin Megl umine .......................................................................................... 50 Sample pre paration .................................................................................... 50 Instrument ation .......................................................................................... 51 Data analys is ............................................................................................. 53 Pharmacokinetic Modeling ............................................................................... 53 Results .................................................................................................................... 54 Phenylbutazone ................................................................................................ 54 Ketoprof en ........................................................................................................ 54 Flunixin Megl umine .......................................................................................... 55 Discussio n .............................................................................................................. 55 3 THE DURATION OF CYCLOOX YGENASE INHIBITION FOR PHENYLBUTAZONE, KETOPROFEN AND FLUNIXIN MEGLUMINE FOLLOWING A SINGLE ADMINISTRA TION ......................................................... 69 Background ............................................................................................................. 69 Materials and Methods ............................................................................................ 70 Animals ............................................................................................................. 70 Standard Training Regimen for the University of Florida Equine Performanc e Lab ........................................................................................... 71 Drug Administration and Sa mple Collec tion ..................................................... 71 Ex vivo COX-1 A ssay ....................................................................................... 73 Ex vivo COX-2 A ssay ....................................................................................... 73 Statistical A nalysis ............................................................................................ 74 Results .................................................................................................................... 74 Ex vivo Inhibition of Ph enylbutaz one ................................................................ 74 Ex vivo Inhibition of Ketoprof en ........................................................................ 75 Discussio n .............................................................................................................. 75 4 UTILIZING A BIOSTATISTICAL APP ROACH TO MODEL COMPARISONS OF PHENYLBUTAZONE, KETOPROFEN, AND FLUNIXIN MEGLUMINE WITH SUPRESSED CONCENTRATIONS OF PGE2 AND TXB2 ..................................... 92 Background ............................................................................................................. 92 Materials and Methods ............................................................................................ 94 Animals ............................................................................................................. 94 Method of A nalysis ........................................................................................... 94 Results .................................................................................................................... 95 Relating Phenylbutazone Concentrat ions to Prostaglandin (PGE) Concentrati ons .............................................................................................. 95

PAGE 7

7 Relating Phenylbutazone Concentr ations to Thromboxane (TXB) Concentrati ons .............................................................................................. 95 Relating Ketoprofen Concentrations to Pr ostaglandin (PGE) Concentrations .. 96 Relating Ketoprofen Concentrations to Thromboxane (TXB) Concentrations .. 96 Relating Flunixin Meglum ine Concentrations to Prostaglandin (PGE) Concentrati ons .............................................................................................. 97 Relating Flunixin Meglum ine Concentrations to Thromboxane (TXB) Concentrati ons .............................................................................................. 97 Example of Model Applicat ion .......................................................................... 98 Discussio n .............................................................................................................. 98 5 CONCLUSION S ................................................................................................... 108 LIST OF REFE RENCES ............................................................................................. 111 BIOGRAPHICAL SK ETCH .......................................................................................... 116

PAGE 8

8 LIST OF TABLES Table page 1-1 Non-steroidal anti-inflamma tory drugs currently availabl e for use in the horse. 39 1-2 Maximum allowable concentrations of NSAIDs in plasma for several equestrian s ports. ............................................................................................... 40 2-1 Plasma concentrations of phenyl butazone as determined by LC/MS/MS analysis .............................................................................................................. 60 2-2 Relevant pharmacokinetic param eters for phenylbutazone following a 2compartmental analysis. ..................................................................................... 62 2-3 Plasma concentrations of ketoprofen as determined by LC/MS/MS analysis ..... 63 2-4 Relevant pharmacokinetic param eters for ketoprofen following a 2compartmental analysis. ..................................................................................... 65 2-5 Plasma concentrations of flunixin meglumine as determined by LC/MS/MS analysis .............................................................................................................. 66 2-6 Relevant pharmacokinetic parameters for flunixin meglumine following a 2compartmental analysis. ..................................................................................... 68 3-1 Ex vivo TXB2 concentrations for phenyl butazone in si x horses .......................... 80 3-2 Ex vivo PGE2 concentrations for phenylbutazone in six horses .......................... 82 3-3 Ex vivo TXB2 concentrations for ket oprofen in si x horses .................................. 84 3-4 Ex vivo PGE2 concentrations for ket oprofen in si x horses .................................. 86 3-5 Ex vivo TXB2 concentrations for flunixin meglumine in 6 horses ........................ 88 3-6 Ex vivo PGE2 concentrations for flunixin meglumine in 6 horses ........................ 90

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Plasma elimination of phenylbutaz one represented by concentration versus time up to 4 h post dr ug administrat ion. .............................................................. 61 2-2 Plasma elimination of phenylbutaz one represented by concentration versus time up to 48 h post dr ug administrat ion. ............................................................ 61 2-3 Plasma elimination of ketoprofen r epresented by concentration versus time up to 4 h post drug administrat ion. ..................................................................... 64 2-4 Plasma elimination of ketoprofen r epresented by concentration versus time up to 24 h post drug adm inistratio n. ................................................................... 64 2-5 Plasma elimination of flunixin meglumine represented by concentration versus time up to 4 h post drug administr ation. .................................................. 67 2-6 Plasma elimination of flunixin meglumine represented by concentration versus time up to 72 h pos t drug administr ation. ................................................ 67 3-1 Mean SD of % inhibition of phenylbutazone on TXB2 concentrations .............. 81 3-2 Mean SD % inhibition of phenylbutazone on PGE2 concentrations. ................ 83 3-3 Mean SD % inhibition of ketoprofen on TXB2 concentra tions .......................... 85 3-4 Mean SD % inhibition of ketoprofen on PGE2 concentra tions ......................... 87 3-5 Mean SD % inhibition of flunixin meglumine on TXB2 concentrations ............. 89 3-6 Mean SD % inhibition of flunixin meglumine on PGE2 concentrations ............. 91 4-1 Raw and Estimated log(PBZ/PGE) versus time. ............................................... 101 4-2 Raw and Estimated log(PBZ /TXB) versus time. ............................................... 102 4-3 Raw and Estimated log(KET O/PGE) versus time. ............................................ 103 4-4 Raw and Estimated log(KET O/TXB) versus time. ............................................ 104 4-5 Raw and Estimated log(FL /PGE) versus time. ................................................. 105 4-6 Raw and Estimated log(FL /TXB) versus time. .................................................. 106

PAGE 10

10 LIST OF EQUATIONS Equation page 4-1 Estimate of the relationship between time and l og(PBZ/PGE ).......................... 101 4-2 Estimated time since adminis tration for log( PBZ/PGE). ................................... 101 4-3 Lower bound of conf idence interv al. ................................................................. 101 4-4 Upper bound of conf idence interv al. ................................................................. 101 4-5 Estimate of the relationship between time and l og(PBZ/TXB) ......................... 102 4-6 Estimated time sinc e administrat ion. ................................................................ 102 4-7 Estimate of the relationship between time and log( KETO/PGE). ...................... 103 4-8 Estimated time sinc e administrat ion. ................................................................ 103 4-9 Estimate of the relationship between time and l og(KETO/TXB) ...................... 104 4-10 Estimated time sinc e administrat ion. ................................................................ 104 4-11 Estimate of the relationship between time and log( FL/PGE). ........................... 105 4-12 Estimated time sinc e administrat ion. ................................................................ 105 4-13 Estimate of the relationship between time and l og(FL/TXB). ............................ 106 4-14 Estimated time sinc e administrat ion. ................................................................ 106 4-15 Example of est. time since adminis tration using log(PBZ /PGE) = 3.5. ............. 107 4-16 Example of determining lower bound for the confidence interval. ..................... 107 4-17 Example of determining upper bound fo r the confidence interval. .................... 107

PAGE 11

11 LIST OF ABBREVIATIONS AQHA American quarter horse association ARCI Association of racing commissioners international ASA Acetylsalicylic acid AUC Area under plasma concentration CINOD Cyclooxygenase inhibiting nitric oxide donors COX 1 & 2 Cyclooxygenase 1 & 2 DTSP Drug testing standards and practices program EIPH Exercise induc ed pulmonary hemorrhage FEI Federation equestre internationale FL Flunixin meglumine KETO Ketoprofen LC/MS/MS Liquid chromatography with tandem mass spectrometry LLOQ Lower Limit of Quantitation LOD Limit of Detection LPS Lipopolysaccharide NSAID Non-steroidal ant i-inflammatory drug PBZ Phenylbutazone PGE2 Prostaglandin E2 QAP Drug testing and quality assurance program TXB2 Thromboxane B2 WADA World anti-doping agency

PAGE 12

12 Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Master of Science PHARMACOKINETIC AND DURATION OF CYCLOOXYGENASE INHIBITION STUDIES OF PHENYLBUTAZONE, KETOPRO FEN AND FLUNIXIN MEGLUMINE IN ATHLETIC THOROUGHBRED HORSES By Jennifer Noelle Hatzel December 2011 Chair: Patrick T. Colahan Major: Veterinary Medical Sciences The Non-steroidal anti-inflammatory dr ugs (NSAIDs) represent a family of therapeutic agents that are one of the most widely administered medications in both human and veterinary medicine. Equine veterinarians hav e been particularly keen of utilizing their anti-inflammatory actions on anything from colic, to endotoxemia or to alleviate pain associated with musculoskeletal injury. The Thoroughbred racing industry has historically represented equine athletes at the peak of their performance abilities and consequently at great risk for injuries a ssociated with this type of athleticism. A variety of NSAIDs such as flunixin m eglumine, phenylbutazone and ketoprofen, have been administered to Thoroughbred racehorses in an effort to alleviate pain and reduce inflammation, ultimately with anticipat ion for continued racing. While NSAID administration for relief of pain is humane and necessary, racing authorities around the world prohibit the unscrupulous use of these drugs to enable a compromised horse to race. The first objective described in these experiments was to characterize the pharmacokinetics of flunixin m eglumine, phenylbutazone and ketoprofen in conditioned

PAGE 13

13 Thoroughbred horses utilizing st ate of the art LC/MS/MS and establish updated pharmacokinetic parameters for each drug. The second objective described in these experiments aimed to identify the extent and duration of cycl ooxygenase inhibition, after a single dose of each drug to evaluate the inhibitory ca pabilities for an active inflammatory event. Finally, the data obtai ned from these experiments was utilized in creating a novel method of determining an estimated time of administration and potentially enhance our knowledge of withdraw al times for each of these drugs. In order to obtain this data 6 Thoroughbred horses (3 mares and 3 geldings) in athletic condition sufficient to work 1 m ile in 2 minutes without undue stresses were included. Each horse in the study was adm inistered a single intravenous dose of one NSAID at the beginning of the week, while re maining in training. Blood samples were collected one hour and immediately prio r to drug administr ation. Following administration of the NSAID, co llections took place at 15 and 30 minutes, 1, 2, 3, 4, 6, 8, 24, 48 and 72 hours. Plasma samples we re obtained following centrifugation and placed in -80 C storage until pharmacokinetic anal ysis by the Florida Racing Laboratory. Plasma and serum was obtained following an appropriate incubation period for each sample and stored at -80 C until further processing. All samples were subjected to PGE2 concentration (plasma) or TXB2 concentration (serum) analysis through the use of a commercially available ELISA kit. The pharmacokinetic data obtained for each NSAID provided an updated overview, along with the ability to detect in creasingly minute concentrations of drug, not previously determined. Challenged by noxio us stimuli, phenylbutaz one was significantly inhibitory (p<0.5) towards TXB2 for 48 h and PGE2 for up to 72 h; ketoprofen

PAGE 14

14 demonstrated inhibitory actions on both medi ators for up to 48 h; and flunixin meglumine significantly inhibited both inflamma tory mediators up to 8 h. By combining the data obtained from these st udies, a statistical model was created to use both the concentration of drug in a plasma sample along with the inhibit ed concentrations of either inflammato ry mediatory (PGE2 or TXB2) to determine an estimated time of administration. The increasing sensitivity of analytical techniques along with the inability to completely correlate a plas ma concentration with clinical efficacy, suggests a need to create novel analytical methods to aid in governing the anti-doping industry in Thoroughbred racing. The current study su ggests a statistical model utilizing both pharmacokinetic and pharmacodynamic parameter s to estimate an ad ministration time. Potentially, this method could be us ed to enhance our existing knowledge regarding withdrawal times and in turn, better educat e veterinarians and trainers towards the consequences of administering a monitored NSAID within a more scientifically established window of time prior to post.

PAGE 15

15 CHAPTER 1 INTRODUCTION The Eicosanoids History The eicosanoid family of metabolites trace their known roots back to the identification of prosta glandin by Swedish scientist, Von Eu ler, in the early 1930s. His observations of smooth-muscle contraction and vasodepressor activities in accessory sex glands led him to the discovery that the ac tive ingredient was a lipid-soluble acid in seminal fluid. He aptly bestowed it the nam e, prostaglandin, to denote the hypothesized origin, the prostate gland1. Concomitantly, two Americ an gynecologists, Kurzrok and Lieb, observed fluctuations in contractility of strips of uterine myometrium which were exposed to semen 2. Additional investigations throughout the 1960s led to the discovery that prostaglandins were simply part of a larger family of acidic lipids sharing a basic 20-carbon unsaturated ca rboxylic acid structure 3. By 1971, Vane, Smith and Willis had demonstrated that aspirin and additional non-steroidal anti-inflammatory drugs (NSAIDs) exert their therapeutic action s by inhibiting prostaglandin biosynthesis, culminating in a shared Nobel Prize in 1982 for their landmark discoveries 4. The family name of the prostaglandins, l eukotrienes and related acidic lipids is derived from the Greek word eikosi translated to “twenty” in a reference to the 20 carbons shared by their precursor essential fatty acids. Initially, the nomenclature of the prostaglandins utilized their chemical st ructure and were designated the letters A-F. More recent scientific advances have led to the discovery of newer compounds such as the cyclic endoperoxides PGG2 and PGH2, prostacyclin, thromboxane A2, leukotrienes, and other eicosanoids 1. Pharmaceutical modulation of these and other enzymes within

PAGE 16

16 the arachidonic acid cascade vi a NSAID administration is one of the most clinically important therapeutic moda lities in veterinary and human medicine today. The Inflammatory Cascade As previously mentioned, the classic pr ostaglandins are organized A-F according to their substituents on the cyclopentane ring of the 20-carbon carboxylic acid (prostanoic acid) of which they all share. They are derived from three different 20carbon polyunsaturated fatty acids: 8, 11, 14-eicosatrienoic acid, containing 3 double bonds (dihomo-linolenic acid); 5, 8, 11, 14-eicosa tetraenoic acid, containing 4 double bonds (arachidonic acid); and 5, 8, 11, 14, 17-eicosapentaenoic acid, containing 5 double bonds (EPA). These essent ial fatty acids yield prostaglandins with one, two or three double bonds remaining on the side chains, respectively and are referred to as mono-, di-, or triunsaturated. This categorization is denoted by a subscript to the letter; for example PGE1, PGE2, and PGE3 5. Unlike other autacoids, the eicosanoids ar e not located locally or stored in tissue pools, rather their synthesis depends upon the release of their fatt y acid precursor, arachidonic acid. An essential fatty acid, arachidonic acid is incorporated into phospholipids of cell membranes through es ter links and can be released by acyl hydrolases like phospholipase A2. Chemical, physical, ho rmonal stimuli or other autocoids activate cytosolic phospholipase A2 by the Ca2+-dependent translocation of group IV cytosolic phospholipase A2. This occurs through the hydroxylation of the sn -2 bond releasing arachidonate which is then rapidly oxygenated through a variety of enzymatic pathways including cytochrome P450, cyclooxygenas es and lipoxygenases 1.

PAGE 17

17 Unesterified arachidonic acid is met abolized through oxygenation and cyclization by the smooth endoplasmic reticulum en zyme, endoperoxide G/H synthase, more commonly referred to as cyclooxygenase (C OX), yielding the cyclic endoperoxide PGG2. The COX enzyme is bifunctional ha ving both COX activity as well as hydroperoxidase (HOX) activity. Unlike li poxygenase (LOX) which has been mainly isolated from the lung, platelets and whit e blood cells, COX is metabolized in many tissue types in a wide variety of mammals5. The cyclic endoperoxide PGG2 is then converted to the related cyclic endoperoxide PGH2 through the HOX activity of the COX enzyme 6. Both of these intermediates are high ly unstable with half-lives of less than 5 minutes and are quickly transformed by cell-s pecific isomerases and synthases into prostaglandins, thromboxane and prostacyclin These prostanoid products are thought to be moved through the cell membrane for re lease via a prostagl andin transporter system 7. The COX enzyme is comprised of three different folding units: an epidermal growth factor-like domain, a membrane-bind ing motif and an enzymatic domain. An entrance channel leading to the active site is formed by three structural helices, through which arachidonic acid may access the interior of the bilayer active site It is this active site, where NSAIDs inhibit the access of ar achidonate to the upper portion of the long hydrophobic channel8. In 1991, two major isoforms of cyclooxygenase encoded by different genes: COX-1 and COX-2 were eluci dated. The first isoform, COX-1, is considered a “housekeeping” prostanoid and is constitutively expressed by cells during basal conditions modulating ph ysiologic functions like gastro -protection of the mucosal epithelial cells and renal blood flow regulation. The second isoform, COX-2, however is

PAGE 18

18 not always detectable during basal conditi ons and is upregulated by inflammatory stimuli released from bacte ria, cytokines discharged from macrophages, shear stress and growth factors, assigning it as the source of prostanoids formed in inflammation and cancer 1,8. Separating these two isoforms into simp listic categories of “good” vs. “evil” is actually a misinterpretation as they func tion coordinately in certain circumstances 9. Although their responsibilities di ffer, both isoforms share a si milar crystal structure as a dimer homotypically inserted into the endopl asmic reticular membrane, along with a 61% similar amino acid identity 10. The most important differ ence in their structure lies within the active sites. The structural alteration found in t he COX-2 isoform is a larger catalytic channel with a side pocket such that arachidonic acid can “squeeze past”. This difference has prompted the pharmaceutical industry to synthesize COX-2 specific inhibitors that reportedly block anti-infla mmatory actions but spare the gastric and antiplatelet side-effects, which will be discussed in more detail 8. In 2002, a splice variant of COX-1 was recognized and designated COX-3 retaining the intron-1 gene sequence which encodes a 30 amino acid sequence inserted into the N-terminal hydrophobi c signal peptide of the enzym e protein. Originally discovered in dog tissues as a centra lly acting cyclooxygenase inhibited by acetaminophen, the highest concentrations of this isoform have been isolated to the brain and nervous tissue. A mechanism for its conversion to an active enzyme has yet to be determined. It was suggested by Chan drasekharan et al. that this isoenzyme, may be a target of NSAID action by so me therapeutic agents like acetaminophen, dipyrone 11 and paracetamol in mice 12 in the central nervous system that produces analgesia and hypothermic properties. Such actions are not affected by COX-2.

PAGE 19

19 The additional prostanoid products of arachidonic acid metabolism, thromboxane and prostacyclin, are also derived from t he actions of enzymes on the intermediate endoperoxide PGH2. Initially isolated from throm bocytes, thromboxane synthase was found to be the enzyme responsible for converting PGH2 into a substance referred to as thromboxane A2 (TxA2) that contains an oxane ring in contrast to the cyclopentane ring found on prostaglandins. Thromboxane A2 plays a vital role in vasoconstriction and is a proaggregate in thrombus formation. It has a half-life of around 30 seconds, and is quickly degraded into the more stable form, thromboxane B2. Prostacyclin or PGI2 is a product of action of the enzyme, prostacyclin synthase, upon PGH2 within vascular tissue. It is structurally different from the other two pr ostanoids by displaying a doublering, as well as demonstrating a short hal f-life of approximatel y 2-3 minutes before conversion to a relatively i nactive but stable 6-keto-PGF1 Prostacyclin exerts its effects as a potent vasodila tor as well as antiaggregat ory action on blood platelets 5. Pharmaceutical Intervention The History and Mechanism of NSAIDs The global therapeutic success, deriv ed from the use of nonsteroidal antiinflammatory drugs (NSAIDs), owes its exis tence to the early herbal folklore of its foundation. An excerpt from the first know n pharmacopoeia, in reference to the willow tree reads: ‘The leaves of the willow being beaten small and drank with a little pepper and wine doe help such as are troubled with the I liaca Passio (colic). The decoction of ye leaves and barck is an excell ent foementation for ye Gout’ 13. With this in mind and in an effort to create a more palatable formula with fewer gastrointestinal side effects for his arthritic father, a chemist at Bayer Labor atories began tinkering with acetylsalicylic acid (ASA) in 1899 and followed soon a fter with the marketing of aspirin14. From

PAGE 20

20 inception, NSAIDs have become one of t he most widely used therapeutics for antiinflammatory disorders in both human and veterinary medicine. Historically described, the classica l signs of inflammation include: calor (warmth), dolor (pain), rubor (redness), tumor (swelling) and functio laesa (loss of function) invoked by a wide variety of noxious stimuli including infe ctions, antibodies and physical trauma. Upon induction of in flammation, prostanoid biosyn thesis is upregulated leading to prostaglandin E2 and prostacyclin production, leadi ng to the subsequent physiological responses such as increased blood flow, vascu lar permeability and leukocyte infiltration. It is through the inhibition of cyclooxygenase, consequently halting th e biosynthesis of prostanoid products, which many NSAIDs ex ert their therapeutic action leading to a decline in cytokine production, leukocyte re cruitment and ultimately inflammation. NSAID administration obtunds pain by s ubduing the cytokines responsible for increasing the sensitivity of nocioceptive pain perception on nerve endings and thereby, reducing hypersensitivity caused by an infla mmatory response. Si milarly, fever is reduced as NSAIDs inhibit PGE2 synthesis, the prostanoid pr imarily responsible for crossing the blood-brain barrier and trigger ing the hypothalamus to elevate body temperature and decrease heat loss14. The NSAIDs available for clinical use are diverse and abundant. They share many common therapeutic goals along with numerous undesirable side effects. This is much in part due to their common hydrophobic st ructure, which allows access to the hydrophobic arachidonate binding channel. The traditional NSAIDs (non COX-2 selective) are generally organic acids with relatively low pKa values and high uptake following oral administration. Once absorbed, they are highly bound to plasma proteins

PAGE 21

21 (95-99%), implying that the free active frac tion of the drug excreted via glomerular filtration or tubular secr etion is very small. Additionally at sites of infl ammation, the pH is lower which encourages many NSAIDs to accumulate in these regions 14. The propensity of NSAIDs to be highly protein bound, causes displacement of other acidic drugs from their binding si tes on plasma proteins if administered concurrently. Dicumarol (Warfarin), as an example, is highly bound to albumin and is displaced in the presence of NSAIDs in a dose-dependent manner. This potentially leads to increased anticoagulant effects as a re sult of elevated conc entrations of free dicumarol in the blood. Similarly, am inoglycoside antibiotic administration in simultaneous administration with an NSAI D can potentially elevate the risk for nephrotoxicity 15. As previously described, activated arachido nic acid enters the active site of both COX-1 and COX-2 enzymes through a hydrophobic channel to initiate the cascade of events observed in inflammation. The mechanism of action for NSAIDs is thought to be similar to that of acetylsalicylic acid to ef fectively inhibit this action, by irreversibly blocking access to the channel through acetylation of serine 530 (COX-1) or serine 516 (COX-2) 10. Once bound, the NSAID COX inhibito rs prevent arachidonic acid from entering the active site of the cyclooxyge nase enzyme, thus prev enting the formation of the characteristic pros tanoid aromatic ring. The therapeutic administrat ion of NSAIDs has o ften been limited by their multisystemic adverse effects, particularly concerning the gastrointestinal system. Upon discovery that COX-2 repres ented the dominant form of inflammatory and cancerous prostanoids, many pharmaceutical laboratorie s pursued identification of NSAIDs that

PAGE 22

22 selectively inhibited this specific isoenzym e. The amino acid conformation along with several subtle differences between the two is oforms accounts for the selectivity of COX1 versus COX-2 inhibitors. The smaller valine residues at sites 434 and 523 allows for a side-pocket to form in COX-2 isofo rms, whereas COX-1 demonstrates larger isoleucine residues that effectively close dow n the entrance to a si de-pocket formation. Optimal orientation with this side-pocket allo ws for COX-2 selective diaryl heterocyclic compounds, to maintain their own bulky side group and form a bond with COX-2 and not COX-1 14. Additionally, COX-2 maintains a charged arginine residue at amino acid 513, whereas COX-1 has an aromatic histidi ne residue in this location. The charged arginine residue interacts with sulphonami de groups often present on COX-2 selective inhibitors, while the imidazole ring of t he histidine residue is unable to interact, demonstrating additional COX sele ctivity. Finally, certain residues present in COX-2 but not COX-1 are able to form a hydrogen bond wit h COX-2 inhibitors, promoting further selectivity 16. Adverse Effects Affi liated with NSAI D Administration Although all NSAIDs provide antipyret ic, analgesic and anti-inflammatory properties (with the exception of acetaminophen which lacks anti-inflammatory activity), relief is not afforded without a cost. Adverse side-effects resulting fr om the use of these drugs have been well documented since the introduction of industrially produced salicylic acid in 1874 and affect both hum an and veterinary patients. Body systems most often affected by NSAID administration include the gastr ointestinal tract, platelets, the renal system, the cardiova scular system, the central nervous system, the female reproductive tract and variable hypersensitivitie s. The constitutive isoform, COX-1, found in blood platelets yields thromboxane A2, of which inhibition leads to a loss of

PAGE 23

23 normal platelet aggregation. The activation of COX-1 also produces antithrombogenic prostacyclin, which is released by the endothel ium along with cytoprot ective prostanoids in the gastric mucosa. Importantly, these eicosanoids inhibit acid secretion by the stomach, enhance mucosal blood flow and promot e secretion of protective mucus in the lumen of the intestine 14. When these cytoprotective actions of prostanoids are inhibited, gastric damage like mucosal ulce ration leading to hemorrhage and potential perforation can occur. This complicati on represents the majo r concern with NSAID administration in human medicine, especia lly regarding the geriatric population of patients. An additional mechanism for N SAID induced ulceration has been proposed to occur due to local irritation through dire ct contact with an orally administered formulation, however this risk is perceived minor when compared with inhibition of the cytoprotectivity actions of COX-1 induced prostanoids 14. COX-1 dependent prostaglandin biosynthesis in the kidney pr imarily occurs in the renal medulla, the ascending loop of Henle and the cortex. Pr ostaglandin synthesis controls several aspects of renal physiology including total renal blood flow, distri bution of renal blood flow, sodium and water reabsorption and r enin release. When these physiologic functions are disrupted by NSAID administrat ion, leading to a decrease in glomerular filtration rate or sodium and water excretion, a variety of clinically relevant effects may ensue including acute renal failure, and hypertension. Novel Classes of Se lective NSAIDs Although our understanding has grown extensively since the discovery of COX-2 in 1991, it is still incomplete. Novel side effects of selective COX-2 NSAIDs, sprung from the selective inhibition, became publically evident between 2001 and 2003. Prior to this public controversy involving the ‘coxib’ drugs, Mukherj ee et al. had raised a

PAGE 24

24 cautionary flag based on a variety of studies indicating an increased risk of cardiovascular events associated with COX-2 inhibitors 17. Subsequent studies determined that the adverse cardiovascu lar complications are associated with prolonged administration of CO X-2 inhibitors including in creased risks of myocardial infarction, destabization of controlled conge stive heart failure and exacerbation of high blood pressure. These complications ultimate ly lead to the withdrawal of some of the COX-2 inhibitors from the market by the federal Food and Drug Administration18. Ongoing pharmacological studies aimed at creating a more gastrotolerant and effective NSAIDs have followed several strat egies. The concept that gastrointestinal safety can be improved by forming nitroso deriv atives of the conventional nonselective COX inhibitors, has led to the introducti on of cyclooxygenase inhibiting nitric oxide donors (CINODs). The in vivo induction of nitric oxide release appears to be gastroprotective, and possibly demonstrates increased anti-inflammatory and analgesic potency. Although promising, to date no drugs in this class have been introduced into the veterinary market. A novel NSAID cl ass, demonstrating dual inhibition has been introduced to both the human and veterinary industries recently. These drugs have been designed to inhibit both COX and LOX (s pecifically 5-LO) that produces a leukotriene from another pathway of the arachidonic acid cascade 13. It is believed that inhibition of this secondary pathway could po tentially increase gastrointestinal safety and create greater analgesic e ffects. An example of this in the veterinary pharmaceutical markets of both Europe and the United States is Tepoxalin (Zubrin ), which is currently advertised for the use in dogs with osteoarthritis. Data has indicated that this drug is more COX-1 inhibitory, but also inhibits LOX activity and has proven

PAGE 25

25 increased gastrointestinal safety 11. Giorgi et al. has recently conducted several studies investigating this drug administered orally to horses, but have found that it is rapidly converted to its acidic metabolite, RWJ-20142. This metabolite has diminished COX-2 and 5-LO, inhibitory activities and thus ulti mately raising concerns about efficacy as a long-term therapeutic option for use in equine patients 19-21. Conventionally, NSAIDs have been organi zed according to their chemical structure dividing them into several classes: derivatives of salicylic acid, propionic acid, acetic acid, enolic acid, fenamic acid, alkanones and diaryl heterocyclic compounds (which represent the COX-2 selective agents or ‘coxibs’). With the introduction of selective COX-2 inhibitors, a new classifi cation system was creat ed in an effort to distinguish agents according to thei r specific selectivity. The selective COX-1 inhibitors including aspirin, are used at low dosages for their inhibition of pl atelet aggregation, and are determined to be more damaging to the gastr ointestinal tract at higher dose rates. The nonselective COX inhibitors have been determined to inhibit both COX-1 and COX2, effectively inhibiting platel et aggregation as well as causi ng significant gastrointestinal and renal disturbances. The selective COX-2 inhibitors have been expected to provide decreased gastrointestinal and renal side-effect s, permitting administration to patients in whom platelet aggregation must be intact. Finally, the highly selective COX-2 inhibitors represent a group of agents currently being test ed, but not yet clinically relevant or available. One further method of classi fying these drugs is though a system based on t1/2, allowing for visual interpretation of t he persistence of plasma concentrations.

PAGE 26

26 NSAID Use in Veterinary Medicine Equine NSAID Administration The NSAID pharmaceutical agents are am ongst the most commonly prescribed therapeutics in equine medicine today. Disco mfort associated with musculoskeletal pain and inflammation represents the most common indication for administration. Phenylbutazone is commonly the drug of c hoice. Another commonly administered NSAID in equine medicine, fl unixin meglumine, is most often used to relieve pain associated with colic, fever, and soft-tissue inflammatory disor ders. Furthermore, flunixin meglumine, has been the primary NSAID used for the tr eatment of clinical signs associated with endotoxemia although other drugs such as ketoprofen and phenylbutazone have been explored to determi ne their anti-endotoxic properties 22. As science and technology have advanced, the number of NSAIDs available for equine practitioners has grown substant ially. Route, frequency, and ease of administration are all consider ations in NSAID selection. The majority of NSAIDs available for use in the horse ar e administered either intravenously (i.v.) or orally (p.o.). Intramuscular administration is not recommended due to the potential necrotizing myositis at the injection site associated with this route of administration by most NSAID preparations 23. Plasma concentration following oral administration of NSAIDs in equine patients depend upon the drug’s tendency to bind to hay and digesta, leading to incomplete immediate absorption followed by a prolonged absorption peak associated with hind gut fermentation 24. This affects the plasma concentration as well as the onset of action for these drugs. There is some clinical evidence regarding the superior efficacy of individual NSAIDs in treating infl ammatory conditions invo lving specific tissue types. Although unsubstantiated, this has l ed to the practice of administering two

PAGE 27

27 different NSAIDs in patients suffering from two different diseases simultaneously. Although there is little proof that efficacy is improved utilizing this method. Evidence from human medicine that s uggests the occurrences of side-effects are increased by simultaneous NSAID administr ation are increased, but minimal studies have been conducted in equine medicine 14. Furthermore, there is not an overwhelming amount of evidence for the selection of one NSAID ov er another and that the decision should be based upon the situation and owner (Table 1-1). As in human medicine, the primary si de effect of NSAID administration, gastrointestinal ulceration, is the most prevalent side effect in horses. The cytoprotective action of prostaglandins in t he gastrointestinal tract of horses preserves the integrity of the gastric mucosa through a variety of different mechanisms that are inhibited with administrati on of NSAIDs. Additionally, the acidity of the drug is hypothesized to directly irritate the gastric mucosa following oral administration, adding to an untoward effect 25. Horses also demonstrate a un ique sensitivity of the right dorsal colon to the ulcerogenic effects of N SAIDs even in the face of moderate dose rates 22. The pathophysiology of this sensitivit y is currently unknown but has been hypothesized to be due to the greater pros taglandin dependency of the blood flow and subsequent vasodilatation in this anatomic region compared to other regions of the colon 26. Although not as notable as the gastrointesti nal effects, renal si de effects resulting from NSAID administration to horses is docum ented in veterinary literature. Renal papillary necrosis most commonly occurs as a sequela to routine doses of NSAID therapy 27 often in dehydrated animals or thos e suffering preexisting impaired renal

PAGE 28

28 perfusion, but does not appear to affect norma l horses. In the ki dney, prostaglandins induce afferent vasodilatation to maintain r enal blood flow and glomerular filtration rate as well as maintain potassium homeosta sis and water and sodium balance. Renal blood flow to the renal pelvis in horses has been hypothesized to be minimal such that even mild dehydration or hypot ension increases the sensitiv ity to ischemic necrosis induced by NSAID administration 27. As previously described, NSAIDs repres ent a family of drugs that are highly bound to plasma proteins. Caution is i ndicated when administering these drugs concurrently with other compound s that are also highly prot ein bound such as additional NSAIDs, sulfonamides, warfar in and gentamicin. Careful monitoring of relevant pharmacokinetic and pharmacodynamic paramet ers is recommended to ensure that the NSAID being used is not competitively inhi biting the binding ability of the other drug and consequently increasing the unbound fraction. Adjustments to dosage should be made in light of this interaction 22. Hepatotoxicity induced by NSAID adminis tration is a potential side effect, reported in a variety of species, but not documented in the horse. The hepatotoxic effects of NSAIDs in other species have been reported to be either idiosyncratic, unpredictable/non-dose related or in trinsic, predictable and dose-related 11. Although the horse does not appear particularly prone to suffering hepatotoxic effects induced by NSAID administration, the potential for this to occur in conjunction with concomitant herbal preparations administ ered by the owner may exist and veterinary consultation should be performed 22.

PAGE 29

29 The effect of NSAID admin istration on coagulatory function in the horse has been minimally investigated but historically, CO X inhibition alters platelet aggregation. Aspirin, despite a short halflife, has been demonstrated to be antithrombotic in horses irreversibly inactivating the aggregatory abilities of COX-1 28. Bleeding time is unaltered by the administration of fluni xin meglumine or phenylbutazone 28, but the effect of any other of the NSAIDs commonly utilized in equine practice is unknown in this regard. With the recent availability of the more COX-2 selective products, an intense interest in the effects of NSAIDs on cartilage as chondroprotective or chondrodestructive, has devel oped. Studies have provided evidence supporting both the protective and destructive properties 29 but more research is necessary to define the effects of NSAIDs in both normal and arthritic joints 22. Drug Doping in the Performance Horse Athletic breakdown on racetracks is unfor tunately a relatively common occurrence in equine competition, causing injury to both horses and riders, tainting the image of horse racing in the public’s eye. Many studies have evaluated different variables potentially leading to a breakdown on t he track but none have drawn specific conclusions. A study was publ ished by Dirikolu et al. in 2008 to evaluate the role of nonsteroidal anti-inflammatory agents in musculoskeletal injuries of racing Thoroughbred horses. The group examined 210 injuries occurring on Kentucky racetracks between January 1, 1995 and December 31st, 1996 and included 161 in the study based on injuries related to the muscu loskeletal system. Plasma samples from the injured horses were assayed for pheny lbutazone, flunixin meglumine and naproxen and were compared with drug concentrations in samples procured from the winner and another horse in the race picked randomly by the stewards of the track. The average

PAGE 30

30 apparent plasma concentration of phenylbutazone in samples from injured horses (5.84 g/mL 0.563), was significantly greater than those from the race winner (4.337 g/mL 0.458) or the randomly selected horse (4.337 g/mL 0.454). None of the concentrations were above an assumed pharmacologically effective level for this study, 7 g/mL. In the same study, plasma conc entrations of flunixin meglumine were significantly higher in plasma from the injured horses (1.632 g/mL 0.158) and in samples from the winning horses (1.067 g/mL 0.078) compared with the samples from the randomly se lected horses (0.695 g/mL 0.069). Most injured horses (81%) had plasma concentrations higher than the a ssumed pharmacologically effective level of 0.1 g/mL. Although no definitive cause and effe ct relationship may be drawn regarding the effects of NSAIDs on racetrack injuries it remains a possibility that the injured horses had preexisting pathology, being masked by the administration of NSAIDs that was accentuated during the race. However, the role of additional risk factors including age, racetrack surface, length of race, gender, training progra m, and preexisting pathologic conditions must be examined further 30. Many states allow the contro lled use of NSAIDs during tr aining to alleviate “sore” horses. The question of NSAID administr ation effects on performance and the ethics behind permitting such treatments may certainl y be raised. Several studies have been conducted examining the effe cts of NSAIDs on exercisi ng horses and potential effects on pharmacokinetics. Soma et al., in concurrence with other studies 31, concluded that age but not training status influenced the disposition of the drug as defined by depression of serum thromboxane levels 32. However, as pain is an indication for the body to spare the use of afflicted tissues competing under the influence of analgesic

PAGE 31

31 drugs may cause further damage and possibly in crease the incidence of breakdowns on the track. Although performance enhancement by drug administration contaminates all aspects of equine sports, the racing industr y has historically suffered the greatest abuse. The sport of horse racing is an ancie nt past-time with the first ever recorded ridden race occurring at the Greek Olympiad in 624 BC 33. With the cultivation of the Thoroughbred in England and the Standardbred in France and the United States throughout the 17th, 18th, and 19th centuries, the foundation for modern horseracing was set and became a globally popular amusem ent. Alongside this entertainment, undoubtedly ignited through human competitiveness, came the desire to utilize chemical substances to improve the prospect of winni ng. However, a stimulatory effect was not always the goal and in fact, the British practice of ‘nobbling’ was on e in which competing horses were subdued through chemical manipulat ion with the purpose of deceiving the betting public 34. With the establishm ent of the English Jockey Club in 1752, came the ‘hay, oats and water’ rule, which has served as the guiding principle throughout the racing world over the past two centuries. Wi th the invention of the hypodermic needle in the 19th century and the development of pure inject able forms of stimul ant drugs, arrived a new era of doping. Monitoring Drug Abuse As advancements within the pharmaceutical industry have substantially increased the number of available doping a gents, the forensic laboratories strive to maintain the number and sensitivity of analytical techniques During the 1940s, very few drugs were detectable and confirmation was difficult usi ng melting point analytic technique, the primary technique available at the time. Chromatography was rediscovered in the

PAGE 32

32 1950s and led to both gas chromatography and qui ckly thereafter, by the thin layer chromatography in the 1960s, allowing for the i dentification of much lower concentration of illegal drugs. The pharmaceutical industr y continually evolved through the design of products boasting higher affinities and greater specificity leading to increased potency and reduced side effects at lower dosages. Throughout the 1980s many of these potent substances began surfacing in test sample s from the racetracks and the forensic laboratories were challenged. With the us e of chromatography/mass spectrometry, the ability to detect minute concentrations of prohibited substances became a reality and has become the cornerstone in drug rule enforcement 33. Ironically, the sensitivity of current analytical techniques have become such that the detection of illegal substances is possible for such a prolonged period fo llowing administration that there may be no significant pharmacologic effect on performanc e from the drug. The debate involving the establishment of threshol d limits for certain substances and how those limits should be set continues to be a ‘hot topic’ in the racing industry. As the need for regulation on racetra cks became more and more evident, the Association of Racing Commissi oners International (ARCI) was created in 1988 to provide uniformity to the rules of racing. Pari-mutuel betting and racing are regulated at the state rather than th e national level in the United Stat es, and currently there are 44 states regulating their own Rules of Racing. The members of the ARCI are politically appointed and function to make recommendati ons on rules but do not serve as a regulatory body. The Drug Testing and Q uality Assurance Program (QAP) was established in 1988 by the ARCI in an effort to monitor the testing laboratories involved in processing racetrack samples and maintain a level of performance standard. In 1995

PAGE 33

33 the QAP was replaced by another progr am called the Drug Testing Standards and Practices Program (DTSP) that currently a ccredits 5 out of 18 l aboratories throughout the U.S. receiving test samples from race tracks. Stewards and commissioners at the state level are responsible for determining the severity of disciplinary action enacted in the face of a post-race positive drug sample Typically the trainer, who is considered the primary insurer of the horse’s fitness to race, receives the brunt end of a punishment, usually in the form of a fine and/or license suspension 33. Although regulation on racetracks is still maintained at the state level, the ARCI has developed a set of guidelines in which commissioners use as a template for their own rules of racing in addition to maintain ing legislation. These procedures have been collected and published in the ‘Model Rules of Flat Racing’ which is a 25 chapter publication containing a wide range of subjects, updated periodically with the most current version published in October of 2010. The eleventh chapter of this publication addresses equine health and medication and covers topics such as: the Uniform Classification Guidelines, penalty recommendati ons, medication restrictions as well as the permitted usage of phenylbutazone and furo semide. The Uniform Classification Guidelines provides a tiered sc hematic, organizing foreign s ubstances according to their pharmacological capabilities of affecting a race along with their appropriateness for use on racehorses in general. The QAP along with an additional ARCI subcommittee, the Drug Penalty and Sample Selection, consider ed input from the Horsemen’s Benevolent and Protective Association, the American A ssociation of Equine Practitioners, the American Horse Council and the Thoroughbred Racing Protective Bureau along with others in order to draft the first edition of these guidelines along with their recommended

PAGE 34

34 disciplinary actions in 1992. The criteria ta ken into consideration in formatting these guidelines included the overall rules of ra cing, the ability of a substance to alter performance, and the pattern of use along wit h a drug’s acceptability to be used as an equine veterinary therapeutic agent. Not only do the guidelines support uniformity involving the pharmacological agents, but they also provide assistance to a commissioner (who may or may not be savvy to this industry) in understanding the significance surrounding a positive finding. A dditionally, positive results may possibly be the result of inattention or variability of a drug administration, and penalization should be applied accordingly 33 As of December 2010, there are over 860 foreign materials listed within the ARCI’s Uniform Classification of Foreign Substance Guidelines. Class one represents the drugs that are among the greatest conc ern due to their pharmacological effect and their potential outcome on a race. Thes e drugs generally do not have acceptable medicinal use on racehorses and include those in the family of opiates, amphetamines, and pemoline. Class two drugs have a high pot ential to affect performance and are not generally utilized in equine medicine consisti ng of psychotropic drugs used by humans, mood-elevating drugs and those used to reduce anxiety. The drugs found in the third class include those that may or may not be generally acceptable for use in the racehorse but demonstrate less potential to alter a race. Included in this class are bronchodilators, which may be used during trai ning, procaine, which often accompanies administration of procaine pen icillin G administration and an tihistamines. The fourth class of substances includes a varied list of drugs more commonly used during training such as: the NSAIDs, steroids, diureti cs (not including furosemide) and other

PAGE 35

35 miscellaneous drugs. Finally, th e fifth class of foreign s ubstances includes those agents that are commonly used in hor ses and are of little interest in affecting performance including anti-ulcer drugs, certain antiall ergic medications and anticoagulant drugs. The ARCI has also developed a penalty recommendat ion scheme located at the end of the Guidelines with advice for NSAID overages as well as for furosemide, which is permitted for the prevention of exerci se-induced pulmonary hemorrhage (EIPH). Finally, there are several drug classes which are considered of no interest in equine regulatory actions and include: antibiotics, sulfonamides, anthelmintics and vitamins 33,35. Every state implements t he guidelines to suit its ow n needs in ways varying from the amount of NSAID administrat ion permitted and the limits of dr ugs to the time prior to post at which furosemide administration is allowed. Several other countries have established unique methods for organizing veterinary medications used on equine racetracks. In Canada, all Pari-Mutuel Wagering is under the jurisdiction of Agriculture Canada, whic h maintains a published list of agents entitled ‘Schedule of Drugs’. The mo st recent edition, published in 2006, presents data from research conducted on Standardbred mares at the Agriculture Canada Equine Drug Evaluation Centre displayed on line as a graph of concentration versus time following administration until the concent ration reaches the detection limit below the analytical method. Trainers and veterinarians can use th is chart as a reference but there are no official recommendations or penalties available. Several of the Provinces will refer to the ARCI Uniform Classification Guidelines for recommendations in penalties when necessary 33,36. The International Agreement on Breeding, Racing and Wagering is a collection of articles, appendices and gui delines published by the International

PAGE 36

36 Federation of Horseracing Authorities to recommend the best practices in significant areas of racing. Approximately fifty c ountries, including the United States, are in agreement upon the most recently updated versi on, published in April 2009. However, not all countries adopt ever y article and for example, the United States excludes sections of Article 6 – Prohibited Substances due to the lack of specificity as well as penalty recommendations within this publication. Owing to the inter national nature of this document and in consideration of the dive rsity of cultures, a classification scheme and/or penalty system would be virt ually impossible to implement 33,37. In examining other equestrian discipli nes, recently the Federation Equestre Internationale (FEI) has adopted the “Clean Spor t” program, which prohibits the use of any NSAIDs (amongst all other foreign prohibi ted substances) immediately prior to and during competition. These new guidelines, adopted in April 2010, mi rror the World AntiDoping Agency (WADA) protocols used in human athletic events and were encouraged to be implemented prior to the 2010 World Equestrian Games in Lexington, Kentucky 38. In the US, the United States Equestrian Federation (formerly the American Horse Shows Association) specifically prohibits the use of more than one NSAID administered at a time and has threshold li mits for individual medications listed in chapter 4 of their rulebook. Additionally, when an NSAID is administered within five days of a competition, a disclosure form is r equired upon entry to the competition 39. The American Quarter Horse Association (AQH A) refers trainers to the Uniform Classification Guidelines recommended by t he ARCI regarding racing Quarterhorses. For showing purposes, the Uniform Classificati on Guidelines are again referenced, but varying threshold limits are listed for indi vidual drugs in addition to the specific

PAGE 37

37 requirement of only one NSAID administered within 24 hours of performance time with the proper disclosur e form (Table 1-2). Summary and Objectives Upon initiation of the inflammatory cascade, a series of enzymatic events converts arachidonic acid release from the cell me mbrane into several endoperoxides, including PGE2 and thromboxane B2. One of these essential enzym es is cyclooxygenase (COX), whereupon the COX-1 isoform is known to be constitutively expressed and vital to tissue homeostasis, but COX-2 is considered inducible and proinflammatory. Both the human and veterinary medical fields have long utilized non-steroidal anti-inflammatory drugs (NSAIDs) in order to inhibit COX, ther eby restricting or reducing the production of prostanoids. While NSAID use is necessary for the relief of pai n, authorities around the world prohibit the unscrupulous use of these drugs to enable a compromised horse to race or perform in any athletic event. Th is has led to the formation of a science specifically responsible for monitoring the doping situatio n in equestrian sports, and subsequently leading to tough decisions rega rding topics such as threshold limits and appropriate punishments. The main focus of this investigation wa s to examine the inflammatory inhibition capabilities along with the decr easing concentrations of three non-steroidal antiinflammatory drugs, commonly used in equine medicine, following a one-time administration. Through several experimen ts, we aimed to supplement the current knowledge of the effects of NSAIDs on at hletic horses and potent ially propose a novel method of determining their e fficacy for commercial use. Specific objectives of the study were as follows:

PAGE 38

38 1. Characterize the pharmacokinetics of phenylbutazone, ke toprofen and flunixin meglumine after intravenous administration to Thoroughbred horses in athletic condition 2. Determine the duration of cycloox ygenase (COX) inhibition of each drug

PAGE 39

39 Table 1-1. Non-steroidal anti-inflammatory drugs currently available for use in the horse. Drug Formula Notes Acetylsalicylic acid (Aspirin) Tablet, powder, paste & gel Not currently U.S. FDA registered for veterinary use but some forms are marketed as if approved. A form in combination with methylpredenisolone is available for use in dogs. Diclofenac (Surpass ) Cream Dipyrone Injectable Only available through compounding in the U.S. but registered in Canada Eltenac (Telzenac ) Injectable Not currently available through the U.S. FDA Firocoxib (Equioxx ) Paste Previcoxx also available for use in dogs in the U.S. Flunixin meglumine (Banamine ) Injectable, paste & granules Ketoprofen (Ketofen ) 10% Injectable solution Available in tablet form & as a 1% injectable soln. for use in dogs & cats in Europe & Canada Meclofenamic acid (Arquel) Granules Available in tablet form for dogs Naproxen (Equiproxen) Granules & injectable Phenylbutazone “bute” (Butazolidine ) Granules, paste & injectable Vedaprofen (Quadrisol) Gel or inject able Only available in Europe & Canada

PAGE 40

40 Table 1-2. Maximum allowable concentrati ons of NSAIDs in plasma for several equestrian sports. NSAID ARCI: Uniform Classification Guidelines (Dec. 2010) FEI (Jan. 2011) USEF (April 2010) A QHA (Dec. 2010) Diclofenac (Surpass ) Prohibited Prohibited 0.005 g/mL 0.005 g/mL Firocoxib (Equioxx ) Prohibited Prohibited 0.240 g/mL 0.240 g/mL Phenylbutazone (Butazolidin ) 2 g/mL Prohibited 15.0 g/mL 15.0 g/mL Flunixin meglumine (Banamine ) 20 ng/mL Prohibited 1.0 g/mL 1.0 g/mL Ketoprofen (Ketofen ) 10 ng/mL Prohibited 0.250 g/mL 40.0 ng/mL Meclofenamic acid (Arquel ) Prohibited Prohibited 2.5 g/mL 2.5 g/mL Naproxen (Naprosyn ) Prohibited Prohibited 40.0 g/mL 40.0 g/mL Eltenac (Telzenac /not yet approved) Prohibited Prohibited 0.1 g/mL 0.1 g/mL Salicylic acid Prohibited Prohibited Prohibited Prohibited

PAGE 41

41 CHAPTER 2 THE PHARMACOKINETICS OF PHENYL BUTAZONE, KETOPROFEN AND FLUNIXIN MEGLUMINE FOLLOWING A SI NGLE ADMINISTRATION Background From the inception of equine athletic co mpetitive events, mankind has attempted to create a “special advantage” in an effort to capture vict ory through the administration of illegal substances. As t he drugs became more complex, the analytical techniques necessary to identify them in blood and urine samples strived to maintain the ability to detect decreasing concentrations. With the advent of complimentary and tandem chromatographic and mass spectrometric techni ques, beginning in the 1980’s, scientists were soon able to detect and quantify physiolog ically irrelevant concentrations of prohibited substances in performance horses33. In conjunction with superior analytical methodologies came the responsibility for un derstanding how certain concentrations of such drugs may cause physiologic changes, which consequently alter the performance ability of horses. Pharmacokinetic and pharmacodynamic (PK-PD) modeling has been utilized in human medicine as a useful technique for over 20 years, and its introduction in veterinary medicine has also proven to be a valuable tool 40. Many of the pharmacodynamic studies performed of NSAIDs on horses have utilized an in vitro technique. However valuable, this method does not take into account the complexity of drug disposition within the whole animal. It was the purpose of this study to evaluate pharmacokinetic data extrapolated from an ex vivo scenario involving three commonly administered NSAIDs to athletic horses, at tempting to mimic the typical time frame surrounding Thoroughbred racing situations.

PAGE 42

42 Materials and Methods Animals Six adult Thoroughbred horses (3 mares and 3 geldings) between the ages of 3 to 10 years, weighing between 495-563 kg, in athletic condition sufficient to work 1 mile in 2 minutes without undue stress were used in this study. General examination of soundness was evaluated prior to inclusion. Routine farriery is performed monthly and dental care (regular floating) is performed annually. Vaccination is administered for tetanus, Eastern Equine Encephalitis (EEE) Western Equine Encephalitis (WEE), West Nile Virus (WNV), Venezuelan Equine Ence phalitis (VEE), influenza and rabies. Treatment with a rotating sc hedule of deworming agents is conducted every 6-8 weeks. The horses were maintained in their typica l paddocks prior to drug administration and for the duration of sample collection. T hey were supplemented with their normal daily ration of sweet feed (Seminole Feed ) and coastal Bermuda hay. This study was approved by and performed in facilities insp ected by the University of Florida Institutional Animal Care and Use Committee (IACUC). Standard Training Regimen for the Universi ty of Florida Equine Performance Lab In an effort to mimic a st ringent exercise program for professional Thoroughbred racehorses, subjects utilized for these studi es were trained three days per week on a high-speed Treadmill (Mustang 2200 ) before and throughout the duration of the study period. The horses were trained for at least 2 months prior to the in itiation of these studies and were administer ed one Standard Exercise Test to verify the level of fitness for inclusion in the st udy. The standard training regi men consists of trotting for 0.6 km at 4.0 m/s, galloping for 2 km at 8 m/s, and trotting for 0.6 km at 4 m/s. On Monday of each week, the training wa s conducted on a horizontal treadmill inclined at

PAGE 43

43 6 from horizontal. Individual records we re kept on each animal indicating date of training, speed, duration of training, and whether samples were collected. Incremental Exercise Test to Exhaustion Following their initial two m onths of training, the horse s were subjected to an Incremental Exercise Test to Exhaustion. The horses were warmed up on the treadmill at 4 m/s for 5 minutes before the start of the test. The horses were then exercised for 1 minute each at 9,10, 11, 12, 13 and 14 m/s until they are unable to maintain the speed of the treadmill with hu mane encouragement including vo cal commands and limited use (3-5 strikes) of a driving whip. Blood samp les were collected via n eedle puncture of the jugular vein for testing of pH and blood lactate, determining whether the horses have met the criteria required for demonstrating an adequate level of exercise stress. Horses demonstrating a blood pH <6.95 and a lact ate concentration >20 mM procured at the time of failure to maintain speed on the treadmill, were included in this study. Standard Condition Test to Verify the Ab ility to Gallop One Mile in Two Minutes The standard training regimen is designed to condition the horses so that they can gallop 1 mile in 2 minutes and recover promptly as indicated by the rate of decrease in their heart rate after the 1-mile gallop. To be included in this study, a horse must periodically demonstrate its ability to meet this ‘mile in 2 minutes’ goal by performing an exercise test at a distance of 1 mile at 13 m/s on the high-speed tr eadmill. The exercise test was conducted after a 4 minute warm up period conducted at 4 m/s. Following completion of the aforement ioned Standard Condition Test, the horse was allowed to recover at 4 m/s for 4 minutes and the heart rate was monito red every 5 minutes. If the heart rate drops below 50 beats/m within 40 minutes of completing the Standard Condition Evaluation, the hor se was considered to have met the fitness goal and was

PAGE 44

44 declared fit to participate in the drug studi es. This challenge was repeated annually and after resumption of training for any horse that may have experienced a significant interruption in its training program. Drug Administration and Sample Collection Phenylbute Injection solution was calculated at the recommended dose of 4.4 mg/kg of body weight for each horse. Ketofen was calculated at the recommended dose of 2.2 mg/kg body weight for each horse and FluMeglumine calculated at the recommended dose of 1.1 mg/kg body weight for each horse. Each individual drug was administered intravenously as a single bolus using a 20-gauge hypodermic needle and syringe via the left jugular vein of each horse on Monday of t he collection week for three weeks. Whole blood samples (7 mL) were collected utilizing venipuncture with a 20gauge vacutainer needle of the jugular vein of each horse immediately prior to drug administration, 15 and 30 min and 1, 2, 3, 4, 6,8, 24, 48 and 72 h following drug administration. Samples destined for t he pharmacokinetic analysis were directly collected into lithium heparinized, vacuumed blood collection tubes and maintained on ice until further processing. The blood samples harvested for pharmacokinetic analysis were immediately centrifuged at 1318 x g (3000 rpm) for 10 min at 4 C. Following centrifugation, the plasma was separated and a liquoted into Cryovial storage tubes at 80 C until drug analysis could be performed. Drug Analysis Chemicals and Reagents Analytical grade drug standards incl uding phenylbutazon e (phenylbutazone-d9) ketoprofen (d3-ketoprofen) and flunixin meglumine (d3-flunixin) were used for this

PAGE 45

45 analysis. Reagent grade formic acid was obtained from ACROS Organics All solvents including acetonitrile, methanol an d dichloromethane were High Pressure Liquid Chromatography (HPLC) gr ade and obtained from Thermo Fisher All of the methods described in further detail below, were validated according to recommendations made by the Food and Drug Administration (FDA). Phenylbutazone Sample preparation All stock standard solutions were pr epared from solid form and dissolved in methanol. All working standard solutions were diluted to the appropriate concentrations in methanol to prepare calibrators in plas ma from 0.005-2.5 ng/mL. Calibrators and positive control samples were prepared from independently prepared stock solutions. Each calibrator and positive control sa mple was prepared from 1 mL of 0.2 M phosphate buffer (pH 3.0) and 0.1 mL of drugfree control horse plasma, and fortified with the appropriate volume of phenylb utazone working standard solution and 25 L of phenylbutazone-d9 working standard solution. The deuterated phenylbutazone analogue was prepared in a working internal standard solution at a concentration of 1.25 ng/ L. The final internal standard concentration was 10 ng/mL of plasma. In duplicate, a 0.1 mL aliquot of each plasma sample was added to 0.1 mL of ultra pure deonized water (with a resistivity gr eater than or equal to 18 mega ohms and organic content less than 10 ppb) and 20 L of 1.25 ng/ L internal standard working solution in 15-mL screw cap disposable, centri fuge tubes. To adjust the pH to 3.5, 75 L of 1 M phosphoric acid (H3PO4) was added followed by 3 mL of methyl tert butyl ether. The tubes were roto-racked for 1015 minutes to allow for thorough mixing and

PAGE 46

46 then centrifuged for 15-20 minutes at 1508 x g (2800 rpm) or until adequate phase separation was achieved. The top aqueous layer was aspirated into waste and a constant volume (2 mL) of the organic layer was transferr ed into a clean 5 mL conical tube. The contents were evapor ated under nitrogen on a TurboVap LV evaporator Sample extracts were then dissolved in 100-150 L of acetonitrile:water (30:70) containing 0.1% (v/v) formic acid and tr ansferred to glass autosampler vials. Instrumentation LC/MS/MS analysis was performed on a Triple Stage Quadrupol e (TSQ) Quantum Ultra Mass spectrometer equipped with a heated electrospray ionization (HESI) source interfaced with a HTC PAL autosampler and Accela LC pump Xcaliber software version 2.0.7 and LCquan version 2.5.6 were used for data acquisition and analysis. Chromatographic separations were achieved with an Acquity™ UPLC HSS T3 (2.1 mm x 50 mm x 1.8 m) column and a 2.1 mm x 5 mm, identically packed, guard column Gradient elution was begun with a mobile phase of 0.1% (v/v) formic acid in water (70%) (Solvent A) and 0.1% (v/v) formic acid in acetoni trile (30%) (Solvent B).The initial mixture, kept constant at a 400 L/min flow rate, was hel d for 0.59 min, then Solvent A was decreased linearly to 0% and Solvent B increased to 100 % over 5.40 minutes. The mobile phase was then returned to the initial conditions for the remaining 0.1 min for a total run time of 6.0 minutes. A divert valve was employed from 0-1.0 min and 2.5-4.0 minutes. The column temperature was isothermal at 45C and a total of 20 L of the sample extract dissolved in 100 L of acetonitrile:water (20:80) containing 0.1% formic acid was injected. Mass spectr al data were acquired in positive ion mode using the HESI and the following MS param eters: ESI spray vo ltage-3000, vaporizer

PAGE 47

47 temperature-198C, sheath gas pressure40, ion sweep gas-0, auxiliary gas pressure15, capillary temperature248C, tube lens offset89, and skimmer offset10. Identification and quantificati on of the analyte were based on selected reaction monitoring (SRM). Compound specific optimiz ation (tuning) of MS /MS parameters was performed before sample via dire ct infusion of 10 ng/L each of the analyte and internal standard dissolved in mobile phase. T uning for phenylbutazone yielded collision energies of 32, 19, and 16 for transitions 309 92, 309 120, and 309 188, respectively. Tuning for phenylbutazoned9 yielded a collision energy of 19 and tube lens offset of 105 for transition 309 120. Data analysis The most abundant ion transmission (188 120) for the analyte was used for quantification. The second and third most a bundant transitions were used as qualifier transitions. All standards, controls, calibrators and samples were prepared in duplicate and peak ion area ratios of the target analyt e and internal standard were calculated for each. Individual values of t he duplicate concentrations were averaged. Calibration was performed using a simple least squar es regression analysis with a 1/Cp weighting factor, where Cp was the nominal plasma concentra tion. Quality control and sample acceptance criteria have been outlined acco rding to the following guidelines and standard operating procedures of the University of Florida Racing Laboratory, Research Division. The requirement is that the %CV for all calibrators, positive controls, and samples must not exceed 10% (15% at t he LLOQ). In addition, for calibrators the difference between the back-calculated conc entration and the nominal concentration must not exceed 10% (15% at the LLOQ). All samples that did not meet such criteria were re-analyzed.

PAGE 48

48 Ketoprofen Sample preparation All stock standard solutions were pr epared from solid form and dissolved in methanol. All working standard solutions were diluted to the appropriate concentrations in methanol to prepare calibrators in pl asma from 0.1-100 ng/mL. Calibrators and positive control samples were prepared from independently prepared stock solutions. Each calibrator and positive control sample were prepared from 1 mL of drug-free control horse plasma, and fortified with the appropriate volume of ketoprofen working standard solution and 20 L of d3-ketoprofen working standard solution. The deuterated ketoprofen analogue was prepared in a worki ng standard solution at a concentration of 0.5 ng/ L. The final deuterated internal standard concentration was 10 ng/mL of plasma. In duplicate, a 1.0 mL aliquot of each plasma sample was added to 1.0 mL of ultra pure de-ionized water (with a resistivity gr eater than or equal to 18 mega ohms and organic content less than 10 ppb) and 20 L of 0.5 ng/ L internal standard working solution in 15-mL screw cap disposable, cent rifuge tubes. To adjust the pH to 3.5, 100 L of 1M phosphoric acid (H3PO4) was added followed by 5 mL of dichloromethane. The tubes were roto-racked for 10 min to allow for thorough mixing and then centrifuged for 16 min at 1508 x g (2800 rpm) or until adequate pha se separation was achieved. The top aqueous layer was aspirated into wast e and a constant volume (3-4 mL) of the organic layer was transferred into a clean 5 mL conical tube. The contents were evaporated under nitrogen on a TurboVap LV evaporator Sample extracts were then dissolved in 100 L of acetonitrile:methanol:water (10:30:60) c ontaining 0.1% formic acid and transferred to glass autosampler vials.

PAGE 49

49 Instrumentation LC/MS/MS analysis was performed on a Triple Stage Quadrupole (TSQ) Quantum Ultra mass spectrometer equipped with a heated electrospray ionization (HESI) source interfaced with a HTC PAL autosampler and Accela LC pump Xcaliber™ software version 2.0.7 and LCquan version 2.5.6 were used for data acquisition and analysis. Chromatographic separations were achiev ed with an Agilent Eclipse XDB-C18 (1 mm x 50 mm x 3.5 m) column and a generic pre-column f ilter. A ternary gradient elution was begun with a mobile phase of 0.1% (v /v) formic acid in water (60%) (Solvent A), 0.1% (v/v) formic acid in methanol (30%) (Solvent B) and 0.1% (v/v) formic acid in acetonitrile (10%) (Solvent C). The init ial mixture, kept constant at a 300 L/min flow rate, was held for 0.75 min, then Solvent A was decreased linearly to 0% and Solvent B increased to 100 % over 4.0 min and held fo r 1.0 min. The mobile phase was then returned to the initial conditions for the remain ing 1.0 min for a total run time of 6.0 min. A divert valve was employed from 0-3. 50 min and 5.50-6.0 min. The column temperature was isothermal at 45C and a total of 20 L of the sample extract dissolved in 100 L of water:methanol:acetonitrile (60:30: 10) containing 0.1% formic acid was injected. Mass spectral data were acquired in positive ion mode using the HESI and the following MS parameters: ESI spray voltage4500, vaporizer temperature-200C, sheath gas pressure60, ion sweep gas-8, auxiliary gas pressure15, capillary temperature398C, tube lens offset89, and skimmer offset10. Identification and quantificati on of the analyte were based on selected reaction monitoring (SRM). Compound specific optimiz ation (tuning) of MS /MS parameters was

PAGE 50

50 performed before analyses via direct infusi on of 10 ng/L each of the analyte and internal standard dissolved in mobile phas e. Tuning for ketoprofen yielded collision energies of 37, 21, and 10 for transitions 255 76, 255 104, and 255 208, respectively. Tuning for d3-ketoprofen yielded a collision energy of 12 and tube lens offset of 98 for transition 258 212. Data analysis The most abundant ion transmission (255 208) for the analyte was used for quantification. The second and third most a bundant transitions were used as qualifier transitions. All standards, controls, calibrators and samples were prepared in duplicate and peak ion area ratios of the target analyt e and internal standard were calculated for each. Individual values of the duplicate concentrations were averaged. Calibration was performed using a simple least squar es regression analysis with a 1/Cp weighting factor, where Cp was the nominal plasma concentra tion. Quality control and sample acceptance criteria have been outlined acco rding to the following guidelines and standard operating procedures of the UF Ra cing Laboratory, Research Division. The requirement is that the %CV for all calibra tors, positive controls, and samples must not exceed 10% (15% at the LLOQ) In addition, for calibrators the difference between the back-calculated concentration and the nominal concentration must not exceed 10% (15% at the LLOQ). All samples that did not meet such criter ia were re-analyzed. Flunixin Meglumine Sample preparation All stock standard solutions were pr epared from solid form and dissolved in methanol. All working standard solutions were diluted to the appropriate concentrations in methanol to prepare calibrators in pl asma from 0.005-50 ng/mL. Calibrators and

PAGE 51

51 positive control samples were prepared from independently prepared stock solutions. Each calibrator and positive control samp le were prepared fr om 1 mL of 0.2 M phosphate buffer (pH 3.0) and 1.0 mL of drugfree control horse plasma, and fortified with the appropriate volume of flunixin wo rking standard solution and 20 L of flunixind3 working standard solution. The deuterated fl unixin analogue was prepared in a working standard solution at a concentration of 0.5 ng/ L. The final deuterated internal standard concentration was 10 ng/mL of plasma. In duplicate, a 1.0 mL aliq uot of each plasma sample was added to 1 mL of phosphate buffer (0.2 M, pH 3.0) and 20 L of 0.5 ng/ L internal standard working solution in 5-mL disposable, centrifuge t ubes. The tubes were centrifuged at 1508 x g (2800 rpm) for 12 min and the buffered plasma samples were subjected to solid phase extraction. Oasis HLB 3-mL columns were sequentially conditioned with 2 mL each of methanol, water, and phosphate buffer. Buffer ed plasma specimens were loaded onto the columns and a positive pressure sufficient to achieve a flow rate of no more than 2 mL per minute was applied. The columns were sequentially washed with 2 mL each of water and methanol:water (10:90). The analyt e was eluted with two 1-mL aliquots of hexanes:ethyl acetate (50:50). The elute was evaporated under nitrogen on a TurboVap LV evaporator Sample extracts were then dissolved in 100 L of acetonitrile:water (10:90) c ontaining 0.1% (v/v) formic acid and transferred to glass autosampler vials. Instrumentation LC/MS/MS analysis was performed on a Triple Stage Quadrupol e (TSQ) Quantum Ultra mass spectrometer equipped with a heated electrospray ionization (HESI)

PAGE 52

52 source interfaced with a HTC PAL autosampler and Accela LC pump Xcaliber™ software version 2.0.7 and LCquan version 2.5.6 were used for data acquisition and analysis. Chromatographic separations were achiev ed with an Acquity™ UPLC HSS T3 (5 mm x 2 mm x 1.8 m) column and a 2.1 mm x 5 mm identically packed, guard column Gradient elution was begun with a mobile phase of 0.1% (v/v) formic acid in water (70%) (Solvent A) and 0.1% (v/v) formic acid in acetonitrile (30%) (Solvent B).The initial mixture, kept constant at a 300 L/min flow rate, was held for 0.5 min, then Solvent A was decreased linearly to 0% and Solvent B increased to 100 % over 2.75 min and held for 0.5 min. The mobile phase was then retu rned to the initial conditions for the remaining 0.75 min for a total run time of 4.0 min. A di vert valve was employed from 01.0 min and 2.5-4.0 min. The column temper ature was isothermal at 45C and a total of 20 L of the sample extract dissolved in 100 L of acetonitrile:water (20:80) containing 0.1% (v/v) formic acid was injected. Mass sp ectral data were acquired in positive ion mode using the HESI and the following MS parameters: ESI spray voltage-3000, vaporizer temperature-198C, sheath gas pressure40, i on sweep gas-0, auxiliary gas pressure15, capillary temperature248C, tube lens offset89, and skimmer offset10. Identification and quantificati on of the analyte were based on selected reaction monitoring (SRM). Compound specific optimiz ation (tuning) of MS /MS parameters was performed before analyses via direct infusi on of 10 ng/L each of the analyte and internal standard dissolved in mobile phas e. Tuning for flunixin yielded collision energies of 48, 38, and 23 for transitions 297 108, 297 264, and 297 279,

PAGE 53

53 respectively. Tuning for d3-flunixin yielded a collision en ergy of 23 and tube lens offset of 133 for transition 300 281. Data analysis The most abundant ion transmission (297 279) for the analyte was used for quantification. The second and third most a bundant transitions were used as qualifier transitions. All standards, controls, calibrators and samples were prepared in duplicate and peak ion area ratios of the target analyt e and internal standard were calculated for each. Individual values of the duplicate concentrations were averaged. Calibration was performed using a simple least squar es regression analysis with a 1/Cp weighting factor, where Cp was the nominal plasma concentra tion. Quality control and sample acceptance criteria have been outlined acco rding to the following guidelines and standard operating procedures of the UF Ra cing Laboratory, Research Division. The requirement is that the %CV for all calibrato rs, positive controls, and samples must not exceed 10% (15% at the LLOQ) In addition, for calibrators the difference between the back-calculated concentration and the nominal concentration must not exceed 10% (15% at the LLOQ). All samples that did not meet such criteria were re-analyzed. Pharmacokinetic Modeling Pharmacokinetic analysis was carried out by non-linear least squares regression analysis determined through the use of a commercially available software package (Phoenix WinNonlin 6.1 NL ME 1.0 BUILD 6.1.0.173 ). The calculations presented were based on equations of pharmacokineti cs described by Gabrielsson and Weiner41. A two-compartmental model wa s determined the best fit for the data including all three drugs in all six horses determined upon visual inspection of the data on a semilogarthmic graph. The compartmental analysis was utilized to determine the area

PAGE 54

54 under the plasma concentration versus ti me curve (AUC), calculated using the trapezoidal rule and extrapolated to infini ty. The values for maximum plasma concentration (Cmax) are reported directly fr om the data. All values are reported as mean SD, unless otherwise noted. Results Phenylbutazone No adverse effects were observed throughout the course of this study. Plasma concentrations of phenylbutazone dete rmined by LC/MS/MS, along with typical statistical values for all six horses in the study were determined (Table 2-1). The plasma concentrations of phenylbutazone versus time for all six horses were plotted for up to 4 hours (Figure 2-1) and 48 hours (Figure 2-2). The plasma concentration for all six horses at the 72 hour time point was below the limit of quantitation (< 10 ng/mL) and not included in the graphical interpretation. Phenylbutazone exhibit ed a long distribution half-life of 2.66 3.40 h, as well as elim ination half life of 117.7 280 h. The median (with range) Cmax determined was 29.8 (16.7 – 43.2) g/mL which occurred at 0.29 0.1 h following administration. Additional relevant pha rmacokinetic parameters are summarized in Table 2-2. Ketoprofen No adverse effects were observed throughout the course of this study. Plasma concentrations of ketoprofen determined by LC/MS/MS, along with typical statistical values for all six horses in the study were determined (Table 2-3). The plasma concentrations of ketoprofen ve rsus time disposition for all six horses were plotted for up to 4 hours (Figure 2-3) and 24 hours (Figure 2-4). The plasma concentration for all

PAGE 55

55 six horses at the 48 and 72 hour time points were below the limit of quantitation (< 0.25 ng/mL) and not included in the graphical interp retation. Ketoprofen exhibited a shorter distribution half-life of 0.461 0.082 h and a longer elimi nation half life of 3.05 0.12 h. The median (with range) Cmax reached was 6.86 (5.41 – 8.13) g/mL, which occurred at 0.25 h following administration. Additional relevant pha rmacokinetic parameters are summarized in Table 2-4. Flunixin Meglumine No adverse effects were observed throughout the course of this study. Plasma concentrations of flunixin meglumine det ermined by LC/MS/MS, along with typical statistical values for all six horses in the study were determined (Table 2-5). The plasma concentrations of flunixin meglumine versus time disposition for all six horses were plotted for up to 4 hours (Figure 2-5) and 72 hours (Figure 2-6). The mean plasma concentration at the 48 and 72 hour time points (0.001 g/mL for each) were still above both the LLOQ at 0.05 ng/mL (0.00005 g/mL) and the LOD 0.01 ng/mL (0.00001 g/mL). Flunixin m eglumine exhibited a short dist ributional half life at 2.05 0.2 h and an elimination half life at 13.1 3.21 h. The median (with range) Cmax reached was 6.51 (4.84-8.06) g/mL, which occurred at 0.25 h follo wing administration Additional relevant pharmacokinetic parameter s are summarized in Table 2-5. Discussion Although a large and diverse family, the pharma cology of most NSAIDs consists of shared characteristics including a comm on mechanism of action leading to their beneficial therapeutic actions. Additiona lly, many of the commonly administered NSAIDs used on horses share similar adverse effects including gastrointestinal and

PAGE 56

56 renal toxicity. Due to their commonality in administration, es pecially in racehorses, they represent a group of drugs with ex tensive metabolic examination 42. Generally, the volume of the central compartment along wit h the volume of distribution are low for NSAIDs, most likely explained by thei r high propensity for protein binding 13. The halflife of most NSAIDs is relatively short wh ich is typical of acid ic drugs, once again, extensively bound to plasma proteins. The pharmacokinetic parameters of phenylbutazone have been particularly scrutinized due to its long-time history as a common analgesic potentially abused in racehorses, although the majority of reports ar e 20-30 years old at this point in time. A more recent study, published in 2010, look ed at the differences in pharmacokinetic parameters of phenylbutazone and dexamethasone in both resting and exercising horses to determine if exercise was a fact or affecting the pharma cokinetic parameters43. Although similar to the current study, their exercise prot ocol aimed to mimic a more rigorous endurance-type training regime whic h lasted for 3 hours and was hypothesized that a shorter amount of exer cise (lasting a few minutes) would likely not alter drug concentrations in plasma or urine. The halflife determined from the horses in this study (3.44 2.42 h) was shorter than the values reported from the af orementioned study for both resting and exercising horses (7.42 1.61 and 7.43 0.99 h respectively). A similar half life was reported in a study conduct ed by Soma et al. with a half life reported at 6.16 h44. It is possible that the larger dose administered in both studies (8 mg/kg) versus the recommended dose used in this st udy (4.4 mg/kg) cr eated a prolonged halflife. The current recommendations published by the RMTC establish a threshold limit of 2 g/mL for racehorses. All six horses in this study demonstrated plasma

PAGE 57

57 concentrations of phenylbutazone below this threshold limit at 24 h although one horse demonstrated a plasma concentration at 24 h close to the threshold limit (1.71 g/mL). The data reported from this study repr esents the most current pharmacokinetic parameters based on a typical on e-time i.v. bolus admin istration of phenylbutazone. Ketoprofen is a propionic acid based N SAID approved for the use in horses for alleviation of pain induced by inflammation and musculoskeletal disorders and by definition, contains a chiral center with two enantiomers. Although marketed and administered as a racemic mixture, it has previously been determined that each enantiomer must be regarded as two distinct drugs13 as stereoselective pharmacokinetics have been report ed. One study in particu lar has already addressed the differences of pharmacokinetics involvin g the two enantiomers, S (+) and R (-), with regards to Ketoprofen45. Since the purpose of this study focused on the clinical effects of each drug, the pharmacokinetics included the total plasma concentration following the administration of the racemate. The published recommended threshold limits for ketoprofen through the RMTC are 0.01 g/mL. All of the horses in this study had plasma concentrations below the allowable thre shold at 24 h, but not at 8 h. The lower limits of quantitation (LLOQ) revealed in this study was 0.25 ng/mL, which is significantly lower than previously reported limits 10 ng/mL, 25 ng/mL and 100 ng/mL4547 respectively. The mean unconjugated plasma ketoprofen concentrations 15 minutes after one administration were 6.94 1.08 g/mL, which is similar to previously reported levels of 8.90 1.90 g/mL46. Sams et al. also reported plasma concentrations of 0.031 0.024 g/mL at the 8 hour time point which is si milar to the concentrations reported in this study 0.014 0.00 g/mL. However, the ability to have a lower limit of quantitation

PAGE 58

58 allowed for detection of plasma concentrati ons of ketoprofen for up to 24 hours following a one-dose administration. The terminal eliminat ion half-life reported in this study (3.05 0.12 h) was longer than previously r eported parameters (1 .60 h and 1.09 h) 45,46, respectively. This could potentially be due to the lower limit of detection in the current study compared with the previous studies allowing for measurement of ketoprofen levels at a slower elimination phase. Flunixin meglumine resides amongst the carboxylic acid nonsteroidal antiinflammatory drugs. This NSAID has been classically used in treating horses for musculoskeletal pain and very commonly admin istered to treat pain associated with a colic incident. Several racing jurisdictions allow this NSAID to be used up to 24 h prior to a race and control potential abuse cases by establishing threshold limits. The current threshold limit recommended by the RMTC is 20 ng/mL when collected following a race. The concentrations determined in this study indicated that all but one horse (1/6) had plasma concentrations of flunixin meglumine below the threshold limit established which demonstrates the variability potential bet ween horses. Previous studies have demonstrated a half life of 2.46 0.97 h31 which differs greatly from what this study revealed with a half life of 13.1 3.21 h. This is most likely due to the differences in the lower limit of quantification (LLOQ) detecta ble between each study. With the advanced analytical techniques utilized in this st udy through LC/MS/MS, the LLOQ established was 0.05 ng/mL versus previously reported limits of 0.1 g/mL. The LLOQ established for this study enabled the detection of fluni xin meglumine up to 72 h, whereas the previously established LLOQ were based upon concentrations only up to 24 h. This led to an estimated half life much greater than pr eviously reported which is an interesting

PAGE 59

59 finding for this drug and a good example of the huge advancements made through the use of sophisticated analytical techniques.

PAGE 60

60 Table 2-1. Plasma concentrations of phenylbutazone as determined by LC/MS/MS an alysis. Concentrations are reported as g/mL. Time(h) Annie Elle Ava Des Slip Tr ue Mean MedianMin Max SD %CVGeomean 0 Below the limit of detection 0.25 43.2 20.1 36.4 26.3 33.3 16.7 29.3 29.8 16.7 43.2 10.1 8.09 27.8 0.5 39.4 15.9 31.4 22.2 26.6 10.3 24.3 24.4 10.3 39.4 10.5 8.43 22.2 1 27.6 11.0 27.5 18.2 17.5 8.83 18.4 17.9 8.83 27.6 7.92 6.34 16.9 2 27.1 10.8 23.2 17.4 16.1 6.47 16.8 16.7 6.47 27.1 7.61 6.09 15.2 3 22.4 9.91 17.5 17.7 10.7 5.61 14.0 14.1 5.61 22.4 6.25 5.00 12.7 4 18.7 8.32 6.29 15.2 8.2 6.32 10.5 8.25 6.29 18.7 5.20 4.16 9.58 6 14.3 5.94 3.94 12.1 7.02 4.41 7.95 6.48 3.94 14.3 4.27 3.42 7.07 8 13.2 3.26 3.54 6.76 6.16 4.32 6.20 5.24 3.26 13.2 3.69 2.95 5.49 24 0.83 0.47 0.75 1.71 0.70 0.49 0.83 0.73 0.47 1.71 0.45 0.36 0.75 48 0.07 0.03 0.06 0.10 0.04 0.02 0.05 0.05 0.02 0.10 0.03 0.02 0.05 72 Below 10.0 ng/mL (Lower limit of Quantitation)

PAGE 61

61 Figure 2-1. Plasma elimination of phenyl butazone represented by concentration versus time up to 4 h post drug administration. Figure 2-2. Plasma elimination of phenyl butazone represented by concentration versus time up to 48 h post drug administration.

PAGE 62

62 Table 2-2. Relevant pharmacokinetic parameters for phenyl butazone following a 2compartmental analysis. Pharmacokinetic Variable Mean SD T1/2alpha (h) 2.35 3.71 T1/2beta (h) 5.73 0.99 Tmax (h) 0.30 0.11 Cmax ( g/mL) 34.3 15.6 k10 (h-1) 0.30 0.20 t1/2k10 (h) 3.44 2.42 k12 (h-1) 1.14 1.98 k21 (h-1) 0.47 0.59 V1 (L/kg) 0.17 0.11 V2 (L/kg) 3.53 7.57 Cl (mL/h/kg) 37.9 14.2 AUC (h* g/mL) 132.5 58.2 T1/2alpha, half-life of distribution; T1/2beta, half-life of elimination; Tmax,, time of maximum concentration; Cmax, maximum concentration; k10, first-order elimination rate constant; t1/2k10, half-life of elimination; k12, rate of transfer from central to peripheral compartment; k21, rate of transfer from peripheral to central compartment; V1, volume of distribution for the central compartment; V2 volume of distribution for peripheral compartment; Cl, systemic clearance; AUC, area under plasma concentration-time curve.

PAGE 63

63 Table 2-3. Plasma concentrations of ketoprofen as determined by LC/MS/MS analysi s. All concentrations are reported as g/mL. Time (h) Annie Tia Ava Des Slip Tr ue Mean MedianMin MaxSD %CV Geomean 0 Below the limit of detection 0.25 8.07 8.13 6.45 7.26 5. 41 6.31 6.94 6.86 5.4 8.1 1.08 15.5 6.86 0.5 3.82 4.14 3.06 4.02 3. 01 3.38 3.57 3.60 3.0 4.1 0.49 13.8 3.54 1 1.50 1.91 1.24 1.66 1.62 1.17 1.52 1.56 1.2 1.9 0.28 18.3 1.49 2 0.32 0.45 0.27 0.48 0.59 0.32 0.40 0.39 0.3 0.6 0.12 30.3 0.39 3 0.13 0.18 0.11 0.17 0.24 0.12 0.16 0.15 0.1 0.2 0.05 29.5 0.15 4 0.07 0.09 0.07 0.08 0.11 0.07 0.08 0.07 0.1 0.1 0.02 20.9 0.08 6 0.03 0.04 0.03 0.03 0.03 0.02 0.03 0.03 0.0 0.0 0.00 13.2 0.03 8 0.02 0.02 0.01 0.01 0.01 0.01 0.01 0.01 0.0 0.0 0.00 23.4 0.01 24 0.00061 0.00069 0.000360.000330. 000370.00036 0.000 0.000 0.0 0.0 0.00 34.8 0.00 48 Below 0.25 ng/mL (Lower limit of Quantitation) 72 Below 0.25 ng/mL (Lower limit of Quantitation)

PAGE 64

64 Figure 2-3. Plasma elimination of ketoprof en represented by concentration versus time up to 4 h post drug administration. Figure 2-4. Plasma elimination of ketoprof en represented by concentration versus time up to 24 h post drug administration.

PAGE 65

65 Table 2-4. Relevant pharmacokinetic parameters for ket oprofen following a 2compartmental analysis. Pharmacokinetic Variable Mean SD T1/2alpha (h) 0.46 0.08 T1/2beta (h) 3.05 0.12 Tmax (h) 0.25 0.00 Cmax ( g/mL) 7.42 1.50 k10 (h-1) 1.43 0.20 t1/2k10 (h) 0.50 0.09 k12 (h-1) 0.02 0.01 k21 (h-1) 0.25 0.01 V1 (L/kg) 307.1 63.3 V2 (L/kg) 112.5 15.9 Cl (mL/h/kg) 430.0 59.7 AUC (h* g/mL) 5.21 0.79 T1/2alpha, half-life of distribution; T1/2beta, half-life of elimination; Tmax,, time of maximum concentration; Cmax, maximum concentration; k10, first-order elimination rate constant; t1/2k10, half-life of elimination; k12, rate of transfer from central to peripheral compartment; k21, rate of transfer from peripheral to central compartment; V1, volume of distribution for central compartment; V2, volume of distribution for peripheral compartment; Cl, systemic clearance; AUC, area under plasma concentration-time curve.

PAGE 66

66 Table 2-5. Plasma concentrati ons of flunixin meglumin e as determined by LC/MS/MS analysis. All values are reported as g/mL. Time (h) Annie Tia Ava Des Slip True Mean MedianMin MaxSD %CV Geomean 0 Below the limit of detection 0.25 8.06 6.77 4.84 6.45 6.57 5.66 6.39 6.51 4.84 8.061.08617.0 6.31 0.5 5.71 6.09 3.25 5.62 5.06 4.86 5.10 5.34 3.26 6.091.01119.8 5.0 1 4.00 4.37 3.02 4.29 3. 63 3.98 3.88 3.99 3.024.360.49612.8 3.85 2 2.29 2.55 2.58 3.08 2. 47 2.74 2.62 2.56 2.293.080.27110.4 2.61 3 1.53 1.86 1.96 2.07 1. 59 1.60 1.77 1.73 1.532.070.22612.8 1.76 4 1.07 1.18 1.28 1.48 1. 05 1.17 1.21 1.17 1.071.480.15913.1 1.20 6 0.57 0.63 0.80 0.82 0. 54 0.61 0.66 0.62 0.540.820.11917.9 0.65 8 0.29 0.29 0.39 0.42 0. 31 0.33 0.34 0.32 0.290.420.05415.9 0.33 24 0.01 0.01 0.01 0.02 0. 01 0.01 0.01 0.01 0.010.020.00429.5 0.01 48 0.00129 0.00103 0.001150.00168 0.000950.000740.001 0.001 0.0 0.000.00028.1 0.001 72 0.00070 0.00049 0.000440.00073 0.000530.000330.001 0.001 0.0 0. 000.00028.3 0.0005

PAGE 67

67 Figure 2-5. Plasma elimination of fluni xin meglumine represented by concentration versus time up to 4 h post drug administration. Figure 2-6. Plasma elimination of fluni xin meglumine represented by concentration versus time up to 72 h post drug administration.

PAGE 68

68 Table 2-6. Relevant pharmacokinetic paramet ers for flunixin megl umine following a 2compartmental analysis. Pharmacokinetic Variable Mean SD T1/2alpha (h) 2.05 0.20 T1/2beta (h) 13.1 3.21 Tmax (h) 0.25 0.00 Cmax ( g/mL) 5.04 0.72 k10 (h-1) 0.33 0.03 t1/2k10 (h) 2.08 0.03 k12 (h-1) 0.004 0.00 k21 (h-1) 0.05 0.01 V1 (L/kg) 0.218 0.03 V2 (L/kg) 0.014 0.003 Cl (mL/h/kg) 72.7 6.33 AUC (h* g/mL) 15.1 1.32 T1/2alpha, half-life of distribution; T1/2beta, half-life of elimination; Tmax,, time of maximum concentration; Cmax, maximum concentration; k10, first-order elimination rate constant; t1/2k10, half-life of elimination; k12, rate of transfer from central to peripheral compartment; k21, rate of transfer from peripheral to central compartment; VI, volume of distribution for central compartment; V2, volume of distribution for peripheral compartment; Cl, systemic clearance; AUC, area under plasma concentration-time curve.

PAGE 69

69 CHAPTER 3 THE DURATION OF CYCLOOXYGENASE INHIBITION FOR PHENYLBUTAZONE, KETOPROFEN AND FLUNIXIN MEGL UMINE FOLLOWING A SINGLE ADMINISTRATION Background Non-steroidal anti-infla mmatory drugs (NSAIDs) represent one of the most commonly administered pharmaceutical therapi es in veterinary me dicine. In modern veterinary medicine the uses of NSAIDs are varied, from lowering body temperature in animals with fever, to relieving respirator y distress in calves and piglets along with controlling postoperative pain in both small and large animals13. Equine veterinarians, in particular, have historically reached into their pharmacy cabinets for an NSAID to treat pain associated with a colic event or musculoskeletal a ilment. The mechanism of action with which NSAIDs exert their therapeut ic effect, involves inhibition of the inflammatory mediator, cycl ooxygenase (COX) which catalyses the conversion of membrane bound arachidonic acid in to prostanoids and thromboxanes22. In 1990, it was discovered that two isoforms of COX existed13. The isoform commonly referred to as COX-1 has been considered the “housekeeping ” isoform, constitutively expressed, and mediates basic functions such as platel et aggregation, renal blood flow regulation and gastric cytoprotection. The other isof orm, COX-2 is considered to be a proinflammation mediator and i nduced by noxious stimuli such as endotoxins, cytokines, stress and injury48. A variety of data has been accumulated regarding the efficacy in veterinary species of both classic and newly synthesized NSAIDs that are CO X isoform-selective and sparing. Researchers have used an asso rtment of mechanisms to compare the effects of different NSAIDs but the com parisons made by different studies have been

PAGE 70

70 difficult to correlate due to variations in study methods. Whole blood assays are well recognized as a convenient way to study the in vitro biochemical efficacy and selectivity of NSAIDs, and are the preferred method48. The advantages of utilizing a whole blood assay are that: (1) they can be used in vitr o in the pre-clinical assessment of COX inhibitors as well as ex vivo; (2) they are sufficiently accurate to estimate potency and selectivity for time-dependent COX inhibitors; (3 ) they compare clinically relevant target cells (platelets and monocytes); and (4) they accommodate the drug to plasma protein binding that occurs in vivo49. It was one of the aims of this study to ev aluate the percent inhibition induced by a one-time administration of three co mmonly utilized NSAIDs (phenylbutazone, ketoprofen, and flunixin megl umine) in athletic Thor oughbred horses, using a whole blood assay. The goal of this study is to define the timeline for an NSAID’s ability to provide inflammatory inhibition. Materials and Methods Animals Six adult Thoroughbred horses (3 mares and 3 geldings) between the ages of 3 to 10 years, weighing between 495-563 kg, in athletic condition sufficient to work 1 mile in 2 minutes without undue stress were used in th is study. Routine farriery was performed monthly and dental care (regular floating) was performed annually. Vaccination for tetanus, Eastern Equine Encephalitis (EEE) Western Equine Encephalitis (WEE), West Nile Virus (WNV), Venezuelan Equine Encep halitis (VEE), influenza and rabies was conducted annually. Treatm ent with a rotating schedule of deworming agents was conducted every 6-8 weeks. The horses were maintained in their typical paddocks for the duration of this study.

PAGE 71

71 The horses were supplemented with daily ration of sweet feed (Seminole Feed ) twice daily and coastal Bermuda hay ad libidium The study protocol was approved by and performed in facilities inspected by the Un iversity of Florida Institutional Animal Care and Use Committee (IACUC). Standard Training Regimen for the Universi ty of Florida Equine Performance Lab Standard training regimen, in cremental test to exhaustion, and standard condition test to verify the ability to gallop one mile in two minutes was previously described in detail previously (Chapter 2). Drug Administration and Sample Collection Each NSAID was administered to each hor se at the manufacture’s recommended dose rate, Phenylbute Injection solution at 4.4 mg/kg of body weight, Ketofen was calculated at 2.2 mg/kg body weight, and FluMeglumine calculated at the recommended dose of 1.1 mg/kg body weig ht. Each drug was administered intravenously as a single dose through a 20-gauge hypodermic needle placed in the right jugular vein. A one week washout period was imposed betwe en administrations. Whole blood samples (7 mL) were co llected via venipuncture with a 20-gauge vacutainer needle of the left jugular vein imm ediately prior to drug ad ministration, and at 15 and 30 minutes and 1, 2, 3, 4, 6, 8, 24, 48 and 72 h following drug administration. Blood samples destined for the m easurement of prostaglandin E2 (PGE2) were collected into partially evac uated blood collection tubes cont aining lithium heparin, and maintained on ice immediately following co llection until processed. Samples to be assayed for thromboxane B2 (TXB) concentrations, were collected into partially

PAGE 72

72 evacuated sterile blood collection tubes c ontaining no anti-coagulant and maintained on ice until processed. The samples obtained for COX-2 i nhibition assay were stimulated for prostaglandin E2 (PGE2) production by incubating with 1 g/mL lipopolysaccharide (LPS, E. coli, serotype 0111:B4 ) within 0.1% bovine serum albumin (Albumin, bovine fraction V ) in phosphate buffered saline (PBS) for 24 hours, as previously described48,50-52. One sample from each horse, was not incubated with LPS, to serve as a negative control. An additional sample, obt ained prior to drug administration, was incubated with LPS to serve as a stimulated positive contro l. Following incubation, meclofenamate sodium salt (100 L mMol solution) was added to each sample tube and gently inverted 3-5 times to mix thor oughly and halt the production of PGE2 53. The samples were then centrifuged at 20,000 x rpm for 10 minutes at 4 C and the plasma separated and pipetted into 1 mL aliquots in to cryovials to be preserved at -80 C until the PGE2 assay was performed. The blood serum samples procured for t he COX-1 inhibition a ssay were allowed to clot for 1 hour at 37 C to allow for maximal produc tion of thromboxane (TXB2) as previously described 48,50-52 and then centrifuged at 20,000 x rpm for 10 minutes at 4 C. The serum was harvested and dispensed into 1 mL aliquots in cryovials for storage at 80 C until assayed for thromboxane (TXB2). A sample obtained prior to drug administration was immediately placed on ic e and centrifuged at 20,000 x rpm for 10 minutes at 4 C and the serum dispensed into 1 mL aliquots for storage at -80 C to serve as a negative control. Another samp le obtained prior to drug administration, incubated for 1 h at 37 C, served as a stimul ated positive control.

PAGE 73

73 Ex vivo COX-1 Assay The aliquots of serum previously frozen fo r storage were allowed to thaw at room temperature. Immediat ely after thawing, 50 L of serum was added to 450 L of methanol and centrifuged at 13, 000 rpm for 5 minutes at 4 C. Following centrifugation, 400 L of the methanol phase was extracted and added to a microfuge tube to be dried down. A CentriVap machine was utilized to dry down the samples overnight at 35 C in order to precipitate the protein component. The resulting precipitant was then reconstituted with 400 L TXB2 enzyme immunoassay (EIA) kit buffer, to yield a final 10fold dilution. The TXB2 levels were then determined according to manufacturer’s instructions for the co mmercially available TXB2 kit (Thromboxane B2 Express EIA Kit ). The baseline and coagulation stim ulated samples were determined and expressed as a mean standard error (SE). Ex vivo COX-2 Assay The aliquots of plasma previously prepar ed and frozen were allo wed to thaw at room temperature. A 100 L sample of plasma was added to 900 L methanol maintaining a 1:10 dilution, vortexed to mi x, and then centrifuged at 13,000 rpm for 5 minutes. Following centrifugation, 850 L of the supernatant was extracted and placed into a microfuge tube. All sample s were positioned withi n the CentriVap and allowed to dry down overnight at 35 C to precipitate the protein. The next day, the resulting precipitant was reconstituted with 170 L PGE2 enzyme immunoassay (EIA) kit buffer to achieve a 1:2 dilution fo r accurate results. The PGE2 levels were then determined according to manuf acturer’s directions for t he commercially available PGE2

PAGE 74

74 kit (Prostaglandin E2 Express EIA Kit ). The baseline and LPS stimulated samples were determined and expressed as a mean standard error (SE). Statistical Analysis The effect of coagulati on and LPS stimulation on TXB2 and PGE2 maximally stimulated samples respectively and was co mpared to inhibition through treatment with phenylbutazone, ketoprofen and flunixin meglumine utilizi ng a two-way ANOVA with a Dunnett’s post hoc test. All calculations were performed through a commercially available software program (SPSS ). Statistical significance was established at p<0.05. Results Ex vivo Inhibition of Phenylbutazone The mean basal values of TXB2 and PGE2 in the six horses was found to be 326.5 232.3 pg/mL and 120.14 67.5 pg/mL respectively. The mean stimulated values of TXB2 and PGE2 were determined to be 40,785.8 10,816.6 pg/mL and 568.7 231.4 pg/mL, respectively. These and all other time points are reported in Tables 3-1 and 3-2. The effects of a clinically relevant one-time dose of phenylbutazone upon the systemic concentrations of TXB2 (COX-1) and PGE2 (COX-2) are depicted in Figures 31 and 3-2, respectively. Statistical anal ysis was performed utilizing a commercially available software package to compare each time point to the maxima lly stimulated time point. This was undertaken to evaluate the inhibition capabilities of phenylbutazone on both inflammatory mediators and determine when a stimulatory effect would reach

PAGE 75

75 maximum levels. Lack of inhibition reached le vels similar to those determined with the maximally stimulated samples at 48 and 72 hours for TXB2 and PGE2 respectively. Ex vivo Inhibition of Ketoprofen The mean basal concentrations of TXB2 and PGE2 in the six horses were determined to be 785.3 1,304.1 pg/mL and 41.5 37.0 pg/mL, respectively. The mean stimulated values of TXB2 and PGE2 were found to be 45,713.4 28,826 pg/mL and 593.2 179.7 pg/mL, respectively. These and a ll other time points are reported in Tables 3-3 and 3-4. The effects of a clinically relevant one-time dose of ketoprofen upon the systemic concentrations of TXB2 (COX-1) and PGE2 (COX-2) are illustrated in Figures 3-3 and 34, respectively. Following statistical analys is, the lack of inhibition as concentrations approached those determined with the maxima lly stimulated sample, occurred at 48 hours for both TXB2 and PGE2. Discussion The present study utilized a whole blood assay, previ ously validated in the horse48 and humans 51, to determine the duration of inhibi tion of two inflammatory mediators and ultimately cyclooxygenases after a one ti me administration of three commonly utilized NSAIDs in athletic Thoroughbred horses. Many ot her studies have previously used a protocol involving a w hole blood assay, primarily to determine selectivity of COX inhibition in the evaluation of novel selective pharmaceuticals 20,48,50. Those studies have determined that TXB2 and PGE2 assayed via this method can be used as markers for COX-1 and COX-2 activity, respectively Utilizing the same methodology, the present study focused on the duration of NSAI D induced suppression for activity of two

PAGE 76

76 inflammatory mediators, TXB2 and PGE2, expressed as a percent of the maximally stimulated positive control sample. The natural clotting process occurri ng in sterile anti-coagulant free blood tubes is a process known to stimulate platel ets to aggregate and activate COX-1. This activation induces the synthesis of the unstable metabolite TXA2 metabolite of arachidonic acid through the COX-1 pathw ay which is quickly hydrolyzed to TXB2 5. Although a rich source of CO X-1, platelets do not normall y produce COX-2, except in clinical conditions associated with a high plat elet regeneration rate. In these situations, the newly released thrombocytes express COX-254. Brideau et al. had previously observed that the levels of TXB2 reached a plateau after the blood was allowed to clot for 60 min and that period has since been used as the standard clotting period in determining TXB2 levels in several studies52. The maximally stimulated concentrations for TXB2 were similar throughout all three assays for phenylbutazone, ketopr ofen, and flunixin meglumine (40785.8 10816.6 pg/mL, 45713.4 28826.3 pg/mL, and 53489.4 24672.8 pg/mL respectively). The slight differences observed reflect the variation of COX activity among individuals and their response to stimulation. The val ues obtained in this study are similar to previously reported values for stimulated levels for TXB2 using a similar assay (40.6 8.5 and 51.6 17.0 ng/mL)20. Both phenylbutazone and ketoprofen suppressed induced levels of TXB2 for about 48 h. After 48 h TXB2 concentrations were statistically indistinguishable from the maxi mally stimulated control values This was identified as the end of inhibitory efficacy. Twenty-four h ours following the administration of flunixin

PAGE 77

77 meglumine, the conc entration of TXB2 was comparable to maximally stimulated levels, indicating the end of the inhibitory efficacy. It has been well documented that the pres ence of lipopolysaccharide (LPS) stimulates cells, especially blood monocytes, to induce CO X-2 to convert arachidonic acid to PGH2 and subsequently PGE2 via PGE synthase48. It has also been determined that human monocytes responded to LPS in a time dependent manner that was slightly enhanced at 4 h post exposure but markedly enhanced at 24 h. Isolated lymphocytes and poly-morphologic neutrophils inc ubated for 24 hours with LPS produced undetectable concentrations of PGE2, indicating that blood m onocytes are the primary circulating producer s of prostaglandins51. Brideau et al. compar ed concentrations of PGE2 production in response to plasma incubation with either LPS or PBS and found that PBS failed to cause any increase in PGE2 activity. In contrast, PGE2 concentrations were comparable to other studies following incubation with LPS52. Earlier research following similar protocols promoted the addition of aspirin to the blood tubes to suppress any undesired COX-1 expression in COX-2 a ssays. Patrignani et al. addressed this issue by incubating human pl asma both in the presence and absence of aspirin following exposure to LPS and f ound no significant difference in the PGE2 production51. Therefore, aspirin use was not incl uded in the COX-2 a ssay used in the current study. Several different concentra tions of LPS have been used in the incubation step of PGE2 stimulation, 10 g/mL and 100 g/mL20,48, respectively. The Brideau study determined from a preliminary LPSdose response experiment that 100 g/mL LPS/mL was required for sustained and consistent PGE2 production in equine blood, but several other studies have used a lower concent ration and obtained similar results.

PAGE 78

78 The PGE2 concentration obtained with maxima l stimulation appeared comparable in all three studies (phenylbutazone: 568.7 231.4 pg/mL, ketoprofen: 593.2 179.7 pg/mL and flunixin m eglumine: 466.6 111.3 pg/mL respectively). The leukocytes of individual horses responded differently to stimulation through LPS, but on average, demonstrated similar production of PGE2. Concentrations of PGE2 in maximally stimulated control samples were comparable to those measured in other studies in which a similar method was carried out (316.6 56.2 pg/mL)20. In this study, phenylbutazone suppressed induc ed concentrations of PGE2 up to 72 h. Thereupon PGE2 production was statistically indistingui shable from the maximally stimulated positive control, indicating an end of inhibitory efficacy at this time point and beyond. One dose of ketoprofen suppressed induced PGE2 concentrations up to 48 h at which time the inhibition efficacy ended and PGE2 concentrations were statistically indistinguishable from those of the maximally stimulated positive control samples. Finally, flunixin meglumine suppressed induced PGE2 concentrations until 24 h after administration, w hereupon induced PGE2 concentrations were statistically indistinguishable from those in the maximally stimulated positive control samples. This study used an ex vivo whole blood assay to determine the inhibitory efficacy and duration following administrat ion of each of three NSAIDs commonly used in equine medicine (phenylbutazone, ketoprofen, and fl unixin meglumine). From this data, phenylbutazone suppresses COX-1 activity un til 48 hours and COX-2 activity through 72 hours following a single admin istration at a 4.4 mg/k g dose rate. Ketoprofen suppresses an induced inflammatory respon se from both COX-1 and COX-2 mediators for up to 48 hours following a single administrati on at a 2.2 mg/kg dose rate. Finally,

PAGE 79

79 flunixin meglumine suppre ssed COX-1 and COX-2 activity through 24 h following a single administration at 1.1 mg/kg dose rate.

PAGE 80

80 Table 3-1. Ex vivo TXB2 concentrations for phenylbutazone in six horses, reported as pg/mL. Time (h) Ava Annie Elle True Slip Des Mean SD Base 695.4 499.1 47.8 195. 4 247.7 273.8 326.5 232.3 Stim mean 30966.1 32763.3 39193.9 60999.037866.542925.9 40785.810816.6 0.25 5319.8 3657.1 631.2 5460.3 496.6 1360.4 2820.9 2290.9 0.5 544.7 652.8 1124.8 489.9 297.0 693.2 633.7 278.2 1 982.7 722.6 2292.5 612.7 508.9 721.7 973.5 665.2 2 1752.5 557.9 2822.4 2354.7 689.0 1350.7 1587.9 901.6 3 2104.9 1940.0 3177.2 4929.3 1638.7 2820.8 2768.5 1203.5 4 2201.2 2719.1 2370.8 3917.5 2786.2 4116.0 3018.5 805.5 6 2764.0 3662.0 8895.5 8256.9 3020.4 6397.0 5499.3 2718.4 8 2383.5 2603.0 24422.54395.6 825.5 2875.5 6250.9 8974.9 24 6951.7 7780.7 37318.922662.114917.310793.8 16737.411604.5 48 31994.2 35951.9 1845.3 57482.0 38976.947540.6 35631.818896.1 72 33047.3 50039.2 283.9 86497.5 51782.960292.5 46990.528795.6

PAGE 81

81 Figure 3-1. Mean SD of % inhibition of phenylbutazone on TXB2 concentrations. *Indicates a significant difference betw een that time point and the maximally stimulated control sample identifying an inhibitory effect.

PAGE 82

82 Table 3-2. Ex vivo PGE2 concentrations for phenylbutazone in six horses, reported as pg/mL. Time (h) Ava Annie Elle True Slip Des Mean SD Base 170.1 143.2 217.2 42. 2 70.5 77.6 120.1 67.5 Stim mean 934.7 740.7 581.8 422.9 383.0 349.0 568.7 231.4 0.25 195.9 77.4 239.7 34.9 131.6 74.2 125.6 79.0 0.5 154.8 86.7 209.0 50.6 118.9 81.2 116.9 57.5 1 224.3 119.0 148.9 100.3 71.0 88.3 125.3 55.4 2 177.1 104.1 430.1 126.4 128.1 90.8 176.1 127.9 3 180.5 124.9 436.5 81.4 108.4 93.9 170.9 134.6 4 163.9 103.4 471.9 97.2 130.6 63.7 171.8 150.8 6 35.17 150.6 457.2 131.2 89.9 106.0 161.7 150.1 8 109.9 145.9 194.5 134.1 85.8 75.6 124.3 43.7 24 216.5 448.0 237.8 289.9 301.5 175.5 278.2 95.4 48 426.4 320.4 260.9 456.8 417.9 288.6 361.8 82.0 72 500.6 391.2 410.9 523.4 324.1 460.2 435.1 74.3

PAGE 83

83 Figure 3-2. Mean SD % inhibition of phenylbutazone on PGE2 concentrations. *Indicates a significant difference betw een that time point and the maximally stimulated control sample identifying an inhibitory effect.

PAGE 84

84 Table 3-3. Ex vivo TXB2 concentrations for ketoprofen in six horses, reported as pg/mL. Time Ava Annie Tia True Slip Des Mean SD Base 3444.2 332.7 324.4 194.2 231.8 184.8 785.3 1304.1 Stim mean 30206.9 29149.3 40204.1103818. 632078.738822.6 45713.428826.3 0.25 182.3 228.3 384.3 300.3 338.6 263.2 282.8 73.7 0.5 320.6 100.2 232.1 217.9 205.6 281.1 226.2 75.3 1 327.7 79.4 160.6 299.5 224. 7 402.8 249.1 118.0 2 349.1 206.1 288.9 305.3 240. 7 481.0 311.9 96.8 3 344.6 156.7 377.5 236.2 301. 7 476.5 315.5 111.6 4 413.0 255.5 533.6 473.0 883. 9 772.6 555.3 233.5 6 614.6 400.4 514.1 1356.4 358. 5 1394.1 773.0 475.2 8 951.7 643.9 465.8 1200.0 579. 6 1278.9 853.3 340.6 24 10394.4 13071.6 14392.654501.8 10360.522763.3 20914.017074.1 48 12413.6 21340.0 40743.663643.9 28862.436333.6 33889.517789.1 72 30628.1 32131.9 50877.179246.1 42606.438409.2 45649.818033.9

PAGE 85

85 Figure 3-3. Mean SD % inhibition of ketoprofen on TXB2 concentrations. *Indicates a significant difference between that ti me point and the ma ximally stimulated control sample identifying an inhibitory effect.

PAGE 86

86 Table 3-4. Ex vivo PGE2 concentrations for ketoprofen in six horses, reported as pg/mL. Time (h) Ava Annie Tia True Slip Des Mean SD Base 50.8 85.0 19.0 60. 0 83.7 3.14 41.5 37.0 Stim mean 768.1 785.5 526.4 706.2 371.9 768.2 593.2 179.7 0.25 79.7 114.3 44.5 136.6 117.3 47.0 86.4 47.6 0.5 33.8 176.0 76.1 97.6 157.3 12.7 85.9 59.7 1 103.7 80.7 79.1 170.5 98.6 56.6 101.2 49.3 2 177.8 80.4 61.1 214.5 97.2 60.9 108.4 72.7 3 166.7 121.5 108.6 168.2 152.29 99.6 132.2 33.3 4 153.2 180.6 162.6 283.6 173.9 53.8 168.5 93.9 6 431.5 396.8 271.3 382.4 216.0 125.3 248.8 107.4 8 112.4 674.4 374.9 351.7 308.6 517.1 388.1 90.3 24 472.7 349.7 591.1 450.5 422.2 478.8 485.7 74.0 48 570.4 595.3 561.2 547.5 506.6 654.5 567.4 62.5 72 494.8 595.0 671.9 481.8 409.7 430.7 498.6 119.5

PAGE 87

87 Figure 3-4. Mean SD % inhibition of ketoprofen on PGE2 concentrations. *Indicates a significant difference between that ti me point and the ma ximally stimulated control sample identifying an inhibitory effect.

PAGE 88

88 Table 3-5. Ex vivo TXB2 concentrations for flunixin megl umine in 6 horses, reported as pg/mL. Time Ava Annie Elle True Slip Des Mean SD Base 256.5 244.5 452.6 67. 1 115.5 505.9 273.7 176.0 Stim 35732.2 47409.4 39558.4100324.3 60418.237493.8 53489.4 24672.8 0.25 294.2 387.1 489.1 324.6 198.8 685.2 396.5 171.3 0.5 397.5 247.9 479.2 246.5 181.3 708.6 376.8 196.4 1 362.4 246.1 387.8 284.4 310.1 908.2 416.5 246.3 2 406.2 488.7 547.4 743.9 527.7 1169.3 647.2 279.1 3 660.0 3746.3 2641.0 1471.4 866.5 1797.3 1863.7 1161.9 4 575.8 728.5 918.8 710.8 628.1 1510.5 845.4 346.3 6 1328.9 971.5 804.6 3179.3 1450.2 2748.2 1747.1 980.5 8 4966.2 7693.3 3318.4 7012.2 2881.5 3434.3 4884.3 2048.9 24 14824.7 28087.7 18908.193986.2 44167. 524832.5 37467.8 29475.5 48 38802.1 56210.5 44318.1109127.972826. 139709.6 60165.7 27200.1 72 27029.9 52232.0 40742.5185738.669186. 662071.4 72833.5 57321.7

PAGE 89

89 Figure 3-5. Mean SD % inhibition of flunixin meglumine on TXB2 concentrations. *Indicates a significant difference betw een that time point and the maximally stimulated control sample identifying an inhibitory effect.

PAGE 90

90 Table 3-6. Ex vivo PGE2 concentrations for flunixin meglumine in 6 horses, reported as pg/mL. Time Ava Annie Elle True Slip Des Mean SD Base 392.87 43.90 69.07 19. 82 31.11 39.51 99.38 144.71 Stim mean 558.16 558. 66 475.96418.79267.26 520.95466.63 111.28 0.25 79.60 118.56 83.04 54. 04 29.50 14.14 63.15 38.35 0.5 86.52 103.96 87.22 34. 92 37.70 68.59 69.82 28.28 1 89.63 103.08 76.05 41. 73 59.33 34.56 67.40 27.00 2 109.53 68.35 5.02 64. 24 47.51 48.66 57.22 34.07 3 83.54 59.30 15.22 60.51 51.71 43.40 52.28 22.57 4 56.44 83.93 26.92 45.40 58.75 95.27 61.12 25.02 6 112.84 98.86 72.73 54. 06 103.7179.00 86.87 22.09 8 123.56 68.72 16.98 80. 53 71.05 80.90 73.62 34.16 24 450.63 590.34 221.23322.56 345.88300.05371.78 130.30 48 508.26 705.72 297.87342.17 340.76306.18416.83 160.98 72 400.40 568.73 255.01303. 44377.57359.40377.42 107.62

PAGE 91

91 Figure 3-6. Mean SD % inhibition of flunixin meglumine on PGE2 concentrations. *Indicates a significant difference betw een that time point and the maximally stimulated control sample ident ifying an inhibitory effect.

PAGE 92

92 CHAPTER 4 UTILIZING A BIOSTATISTICAL APPRO ACH TO MODEL COMPARISONS OF PHENYLBUTAZONE, KETOPROFEN, AND FLUNIXIN MEGLUMINE WITH SUPRESSED CONCENTRATIONS OF PGE2 AND TXB2 Background In addition to the anabolic-androgenic st eroids and sedatives/tranquilizers, the non-steroidal anti-inflammatory drugs (NSAIDs ) represent one of the most extensively studied class of drugs in horses42. As previously discussed, the Association of Racing Commissioners International plac e most NSAIDs within the 4th tier of drug classification, however, some (novel COX-2 selective drugs for example) may be promoted to the 2nd or 3rd tiers if they are not licensed for use in horses55. The decisions surrounding withdrawal times and subsequent penalties regarding the use of NSAIDs on performance horses remains very controversial today. Often control is based upon which drug has been applied, whether it was administer ed prior to or during the competition, and whether or not the drug has been declared. Most of these decisions are based upon pharmacokinetic data that has been accumulated through numerous studies to evaluate the characteristic patte rn of a drug’s absorpt ion, distribution, metabolism, and excretion as a function of ti me. The scientific methodology utilized has improved significantly and curr ently liquid-chromatography linked to mass spectrometry is the method of choice for detection of dr ug concentrations. Historically, equine urine had been the biological sample of choice for detection purposes, because it is easily obtained under post-racing situations and because urine contains a higher concentration of drug metabolites for a longer time than plas ma. However, plasma has gained popularity, especially in the Unit ed States, because it is easily and quickly collected during training or pre-race time points, less sample preparation is required

PAGE 93

93 compared to urine and blood drug conc entrations correspond more closely to pharmacological effects42. Additionally, the once elusiv e minute amounts of drug within plasma are now measurable using advanced screening techniques. More recently, researchers interested in better understanding the pharmacologic effects a drug has on a horse have employed the technique of pharmacokinetic/pharmacodynamic (PK/PD) m odeling. These models have proven to be particularly useful in establishing compet ition drug thresholds, particularly because drug effects do not always correlate with plasma concentrations56. Pharmacodynamics describe the changes of a measurable response of a drug over time. Several different models have been used with equine subjects including heart ra te monitoring, spontaneous locomotor activity, head ptosis and tissue inflammation. The last example, tissue inflammation, has been used in several experiments to examine the pathophysiolgical variable to correlate with drug concentration in PK/PD studies of NSAIDs24. The model in Maihto’s study, used subcutaneously implanted tissue cages and carrageenan-soaked sponges to monitor the penetration of phenylbutazone and eicosanoid concentrations in inflamed tissue. Similarly, our study used an ex vivo approach to stimulate inflammatory mediator s and examined the inhibitory effect of three commonly administered NSAIDs in equine medicine (phenyl butazone, ketoprofen and flunixin meglumine). Finally, employment of a modern biostatistical analysis to correlate the concentration of each NSAID with it s inhibitory effect on both prostaglandin E2 and thromboxane B2, in the stimulated ex vivo inflammatory event represented the final step. The ultimate use of our drug c oncentration to response model is to enable

PAGE 94

94 extrapolation of an estimated administration time by dete rmining both the inhibitory effects and drug concentrations. Materials and Methods Animals Plasma drug concentration and COX-1 and -2 inhibition data was obtained from the six adult Thoroughbred horses (3 mares and 3 geldings) between the ages of 3 to 10 years, weighing between 495-563 kg, in athletic condition sufficient to work 1 mile in 2 minutes without undue stress previously descr ibed in this study. The pharmacokinetic and pharmacodynamic data obtained and described previously was used in the statistical analysis and development of a biostatistical model. Method of Analysis For each horse in the data set, we used either the prostaglandin (PGE) or thromboxane (TXB) and phenylbutazone (PBZ ), ketoprofen (KETO) and flunixin meglumine (FL) concentrations from 0.25 h ours to 72 hours. As our outcome variable, we modeled the natural log of the ratio of each drug concentration to PGE or TXB concentration. To determine the relationshi p between time and log( drug/PGE or TXB), we used general estimating equations (SAS PROC GENMOD) employing a commercially available statistical software package (SAS ), and treating each horse as a random effect and modeling the within-hor se variance between time points with an autoregressive variance-covariance matrix st ructure. Since the relationship between time and log(drug/PGE or TXB) was non-linear, we included a qua dratic term in time to improve model fit. As the number of subjects was few and the object was to determine a relationship that could be used to estimate the time since drug administration from observed

PAGE 95

95 concentrations for all horses, we did not try to find the “best” model fit in order to avoid over fitting the data (creating a model that was a very good estimator for horses in the data set, but poor for other horses). Henc e, we added only a quadr atic modeling term even though in this analysis and most of th ose following, a cubic term or a more complicated non-linear model would have improv ed the fit of the model to the data. The relationships presented here ar e approximate, and valid conf idence intervals that can easily be calculated for any given observation of the two compounds in question. The University of Florida Biostatistical Consul ting Lab, College of Public Health and Health Professions, performed all statistical analyses. Results Relating Phenylbutazone Conc entrations to Prostagla ndin (PGE) Concentrations The model created provided an estimate of the relationship between time and log(PBZ/PGE) illustrated in Equation 4-1. The P-values for all coefficients in this model were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (5.023, 5.490), Time: (-0.230, -0.171), Time*time: (0.0006, 0.0014). In order to use this model, given an observed value of log(PBZ/PGE), one may solve the quadratic formula for an estimated time (Equation 4-2). The smaller of the two solutions is the model estimate of the time since administration. To get confidence limits for this estimate, the same formula can be used, but by substituting the lower and upper limits for the coefficients (Equations 4-3 & 4-4). Alternatively, an estimation of time can be extrapolated from the plot illustrated (Figure 4-1). Relating Phenylbutazone Concentrations to Thromboxane (TXB) Concentrations The model created here provided an estimate of the relationship between time and log(PBZ/TXB) illustrated in Equation 4-5. The P-values for all coefficients in the model

PAGE 96

96 were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (2.448, 3.208), Time: (-0.341, -0.245), Time*time: (0.0013, 0.0027). In order to use this model, giv en an observed value of log(PBZ/TXB), one may solve the quadratic formula for an estimat ed time (Equation 4-6). To get confidence limits for this estimate, the same formula can be used, however, appropriate coefficients would need to be used. Alternatively, an es timation of time can be extrapolated from the plot illustrated (Figure 4-2). Relating Ketoprofen Concentrations to Prostaglandin (PGE) Concentrations The model created here provided an estimate of the relationship between time and log(KETO/PGE) illustrated in Equation 4-7. The P-values for all coefficients in the model were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (-3.191, -2.167), Time: (-1.327, -1.115), Time*time: (0.0272, 0.0351). In order to use this model, given an observed value of log(KETO/PGE), one may solve the quadratic formula for an estimated time (Equation 4-8). To calculate confidence limits for this estimate, the same formula can be used however, appropriate coefficients would need to be used. Alternatively, an estimation of time can be extrapolated from the plot illustrated (Figure 4-3). Relating Ketoprofen Concentrations to Thromboxane (TXB) Concentrations The model created here provided an estimate of the relationship between time and log(KETO/TXB) illustrated in Equa tion 4-9. The P-values for all coefficients in the model were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (-4.186, -3.568), Time: (-1.279, -1.077), Time*time: (0.0215, 0.0295). In order to use this model, gi ven an observed value of log(KETO/TXB), one may solve the quadratic formula for an estima ted time (Equation 4-10). To calculate

PAGE 97

97 confidence limits for this estimate, the sa me formula can be used however, appropriate coefficients would need to be used. Alter natively, an estimation of time can be extrapolated from the plot illustrated (Figure 4-4). Relating Flunixin Meglumine Concentrations to Prostaglandin (PGE) Concentrations The model created here provided an estimate of the relationship between time and log(FL/PGE) illustrated in Equation 4-11. The P-values for all coefficients in the model were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (-2.630, -2.006), Time: (-0.391, -0.372), Time*time: (0.0031, 0.0033). In order to use this model, given an observed value of log(FL/PGE), one may solve the quadratic formula for an estimat ed time (Equation 4-12). To calculate confidence limits for this estimate, the sa me formula can be used however, appropriate coefficients would need to be used. Alter natively, an estimation of time can be extrapolated from the plot illustrated (Figure 4-5). Relating Flunixin Meglumine Concentrations to Thromboxane (TXB) Concentrations The model created here provided an estimate of the relationship between time and log(FL/TXB) illustrated in Equation 4-13. The P-values for all coefficients in the model were highly significant (p<0.0001 for all). A 95% confidence interval for the coefficients were as follows: Intercept: (-4.844, -4.287), Time: (-0.550, -0.479), Time*time: (0.0040, 0.0050). In order to use this model, giv en an observed value of log(FL/TXB), one may solve the quadratic formula for an estimat ed time (Equation 4-14). To calculate confidence limits for this estimate, the sa me formula can be used however, appropriate coefficients would need to be used. Alter natively, an estimation of time can be extrapolated from the plot illustrated (Figure 4-6).

PAGE 98

98 Example of Model Application Suppose you observe log(PBZ/PGE) is equival ent to 3.5. Then the estimated time since administration could be solved by appl ication of Equation 4-15, obtaining the answer of 9.2 hours. Then, maintaining a starting point of 3.5 for the log(PBZ/PGE) value and the lower and upper bounds were applied to determine the confidence intervals (Equations 4-16 & 4-17), an estimation of time could be 95% confident that the drug administration occurr ed between 6.7 and 13 hours ago. Discussion Through the use of an ex vivo model of inflammation that employed a whole blood assay and advanced liquid chromatography wit h tandem mass spectrometry analytical techniques, this study examined both the pharmacokinetic and pharmacodynamic parameters of three NSAIDs (phenylbutazone, ket oprofen, and flunixin meglumine) commonly administered to athletic Thoroughbr ed horses. Previous studies have used similar methodologies to determine potency ratio for COX inhibition 48,50. Most of those studies initially focused on assay tec hnique validation and t hen evaluated the COX selectivity of classic and novel NSAIDs. The current study developed biostatistical formulas to interpret the data and provide enhanced estimations associated with the clinical dose of an NSAID using both pharmacokinetic and pharmacodynamic data. Studies measuring the level of inhibition of a single mediator to predict dosage are of value from a mechanistic perspective but do not necessarily provide information on the clinical response. The degree of prostaglandin inhibition is assumed to correspond to the intensity of clinical responses but exac t correlations are not known. Also, various clinical responses like analgesia, antipyresis and/or resolution of swelling may be induced at different levels of prostaglandin suppression13. The science of relating

PAGE 99

99 plasma concentrations and inflammatory medi ator inhibition to specific effects is incompletely described and large deficienc ies in our understanding of the clinical efficacy of NSAIDs remain. Recently, regulatory authorities have been faced with the challenge of determining the concentration of NSAIDs in pos t race samples that constitute a violation of racing regulations. The appr oach of reporting all drugs detected by very sensitive analytical techniques can be problemat ic because very low but detectable concentrations likely have no effect on raci ng performance. To simplify enforcement and provide scientifically based regulatory de cisions, selected cut-off values have been established through pharmacokinetic and pharmacodynamic data in conjunction with understanding what is acc eptable or unacceptable to regulatory authorities57. The definition of a “limit”, “t hreshold” or “cutoff” is based on any drug or metabolite concentration in a biological fl uid from a participant in a regulated event. When referring to horse racing, concentrations greater than t he stipulated limit induce regulatory action, while those below, do not. In 1999, Tobin et al. proposed the formation of a database containing limitations and withdrawal time data for drugs commonly used in equine medicine, but restricted it to include those medications listed by the AAEP as therapeutic agents. The purpose of this database would be to determine the “highest no effect dose” of a particular drug and ultimate ly the relevant “no effect point” in plasma or urine concentration. Difficulties in es tablishing this database include dose and route of administration, along with the sensitivity of t he analytical methodology used58. The lack of a database has long been deemed a worl dwide problem in the horse racing industry and several different countries have ta ken a variety of approaches to alleviate

PAGE 100

100 the ambiguity associated wit h drug testing. This st udy excluded pre-established “threshold limits” but sought to describe pl asma concentrations and inhibitory efficacy through measurement of two inflammatory mediators (PGE and TXB) following administration. This approach focused on the determination of an appropriate pre-race administration time to meet regulatory needs and was suppor ted by drug effect data through inhibition of inflammato ry mediators. The inhibiti on of inflammatory mediators could be used in conjunction with the rout ine screening process to retrospectively determine time of administration. This information enables humane considerations by providing the time required following NSAID administered in the course of proper veterinary care, to horses in training before participation in regulatory agency sanctioned events. In addition, regulatory agencies gain additional information about the duration and magnitude of t he effects of NSAID administration and their correlation with plasma drug concentration. Potentially, these estimat ed formulas of administration time could be established for a variety of commonly administered foreign substances and provide accentuated information employed by regulatory agencies.

PAGE 101

101 Equation 4-1. Estimate of the rela tionship between time and log(PBZ/PGE). Log(PBZ/PGE)=5.257 – 0.201*T + 0.001*T^2, where T=time (in hours) Equation 4-2. Estimated time sinc e administration for log(PBZ/PGE). T = [0.201 – sqrt(0.201^2 – 4*0.001*(5. 257 – observed log(PBZ/PGE)))]/(2*.001) Equation 4-3. Lower bound of confidence interval. CI = [0.230 – sqrt(0.230^2 – 4*0.0006*(5.0 23 observed log(PBZ/PGE)))]/()]/(2*.0006) Equation 4-4. Upper bound of confidence interval. CI = [0.171 – sqrt(0.171^2 – 4*.0014*(5.490 observed log(PBZ/PGE)))]/()]/(2*.0014) Figure 4-1. Raw and Estimated log(PBZ/PGE) versus time.

PAGE 102

102 Equation 4-5. Estimate of the relationship betw een time and log(PBZ/TXB). Log(PBZ/TBX)=2.828 – 0.293*T + 0.0020*T^ 2, where T=time (in hours) Equation 4-6. Estimated ti me since administration. T = [0.293 – sqrt(0.293^2 – 4*0.002*(2.828 observed log(PBZ/TBX)))]/(2*.002) Figure 4-2. Raw and Estimated log(PBZ/TXB) versus time.

PAGE 103

103 Equation 4-7. Estimate of the relati onship between time and log(KETO/PGE). Log(KETO/PGE)=-2.679 – 1.221*T + 0.0312*T^ 2, where T=time (in hours) Equation 4-8. Estimated ti me since administration. T = [1.221– sqrt(1.221^2 – 4*0.0312*(-2.679 observed log(KETO/PGE)))]/(2*.0312) Figure 4-3. Raw and Estimated log(KETO/PGE) versus time.

PAGE 104

104 Equation 4-9. Estimate of the relationship betw een time and log(KETO/TXB). Log(KETO/TXB)=-3.877 – 1.178*T + 0.0255*T^ 2, where T=time (in hours) Equation 4-10. Estimated ti me since administration. T = [0.1.178 – sqrt(1.178^2 – 4*0.0255*(-3 .877 observed log(KETO/TXB)))]/(2*.0255) Figure 4-4. Raw and Estimated log(KETO/TXB) versus time.

PAGE 105

105 Equation 4-11. Estimate of the relati onship between time and log(FL/PGE). Log(FL/PGE)=-2.318 – 0.381*T + 0.0032* T^2, where T=time (in hours) Equation 4-12. Estimated ti me since administration. T = [0.381 – sqrt(0.381^2 – 4*0.0032*(-2. 318 observed log(FL /PGE)))]/(2*.0032) Figure 4-5. Raw and Estimated log(FL/PGE) ve rsus time.

PAGE 106

106 Equation 4-13. Estimate of the rela tionship between time and log(FL/TXB). Log(FL/TXB)=-4.565 – 0.515*T + 0.0045* T^2, where T=time (in hours) Equation 4-14. Estimated ti me since administration. T = [0.515 – sqrt(0.515^2 – 4*0.0045*(-4.565 observed log(FL/TXB)))]/(2*.0045) Figure 4-6. Raw and Estimated log(FL/TXB) versus time.

PAGE 107

107 Equation 4-15. Example of est. time sinc e administration using log(PBZ/PGE) = 3.5. T = [0.201 – sqrt(0.201^2 – 4*0.001*(5.257 observed log(PBZ/PGE)))]/(2*.001) T = [0.201 – sqrt(0.201^2 – 4*0.001* (5.257-3.5))]/(2*.001) = 9.2 hours Equation 4-16. Example of determining lower bound for the confidence interval. CI = [0.230 – sqrt(0.230^2 – 4*0.0006*(5.0 23 observed log(PBZ/PGE)))]/()]/(2*.0006) CI = [0.230 – sqrt(0.230^2 – 4*0.0006*(5.023 – 3.5))]/()]/(2*.0 006) = 6.7 hours Equation 4-17. Example of determining upper bound for the confidence interval. CI = [0.171 – sqrt(0.171^2 – 4*.0014*(5.490 observed log(PBZ/PGE)))]/()]/(2*.0014) CI = [0.171 – sqrt(0.171^2 – 4*.0014*(5. 490 – 3.5))]/()]/(2*.0014) = 13.0 hours

PAGE 108

108 CHAPTER 5 CONCLUSIONS The overall aim of this study was to enhance our knowledge of the pharmacological effects of three NSAIDs commonly administered to athletic Thoroughbred horses. The pharmacokinet ics of phenylbutazone, ketoprofen and flunixin meglumine were est ablished following a single intravenous administration. A variety of NSAIDs have historically been administered to Thoroughbred racehorses in an effort to alleviate pain and reduce infl ammation, ultimately with anticipation for continued racing. As mentioned, NSAID admini stration for relief of pain is humane and necessary, however racing authorities around t he world prohibit the unscrupulous use of these drugs to enable a compromised horse to race. The drug testing industry has grown substantially for racing Thoroughbreds and the sophisticated analytical methods currently in use are capable of detecting minute amounts of drug concentrations in plasma for a long time after administration. This creates a challenging situation for not only track stewards faced with a positive samp le, but for the veterinarians maintaining the health of these animals making the decis ions regarding drug administration. The withdrawal times currently available are simple recommendations for veterinarians based upon published detection times but more dependent upon clinical judgment. The study described provided an up to date analysis of the pharmacokinetic parameters of three commonly administered NSAIDs (phenylbutazone, ketoprofen and flunixin meglumine) through t he use of advanced LC/MS/MS detection methodology. Furthermore, this study aimed to dete rmine the duration of cyclooxygenase (COX) inhibition of each drug and determine a comparable relationship (if any) between the two. Ultimately, the knowledge obtained was used to create a statistical model

PAGE 109

109 representing the relationship between plasma concentration and inhibitory efficacy of each NSAID. This model provided a novel method for determining an estimated time of administration for each drug. The information identified through this study paves the way for a variety of future directions. The effects on inhibition of inflammation were determined through measurement of the concentrations of two inflammatory mediators (PGE2 and TXB) but the effects on pain through the use of these NSAIDs is st ill undetermined. A similar study utilizing a pain model would be beneficial in determining analgesic effects of these NSAIDs. Additionally, this study used a onetime bolus approach to dose administration and it is well known that many of these drugs are administe red over a longer period of time prior to racing. Therefore, ext ending this study to include a multi-dose administration protocol could enhance the current knowledge regarding the pharmacokinetics involving a more “real -life” method of administration. Importantly, the knowledge gained through this study could potentially lead to more accurate withdrawal times based upon plasma concentrations and inhibitory effects derived from administered NSAIDs. In the future, veterinarians faced with the responsibility for educating their clients would be able to access information regarding the use of a certain drug in a racing Thor oughbred and determine the length of efficacy. This method of providing a more customized withdrawal time for a drug could potentially discourage the use of drugs in a horse too compromised to race. The analytical methodology for detecting doping agents in T horoughbred race horses, for the purpose of monitoring and control, will undoubtedly continue to increase in their level of sophistication. A novel tec hnique for addressing this issue necessitates evaluation to

PAGE 110

110 protect veterinarians, trainers and most impo rtantly, the welfare of Thoroughbreds and the sport of horseracing.

PAGE 111

111 LIST OF REFERENCES 1. Smyth EM, Burke A, FitzGerald GA. Li pid-Derived Autacoids: Eicosanoids and Platelet-Activating Factor. In: Br unton LL, Lazo JS, Parker KL, eds. The Pharmacological Ba sis of Therapeutics 11st ed. New York: McGraw Hill; 2006:653-670. 2. Kurzrok R, Lieb CC. Biochemical studi es of human semen: II. The action of semen on the human uterus. In: Proceedings of the Society for Experimental Biology and Medicine .Vol 28. 1930:268-272. 3. Bergstrom S, Samuelsson B. The prostaglandins. Endeavour 1968;27(102):109-113. 4. Vane JR. Inhibition of prostaglandin syn thesis as a mechanism of action for aspirin-like drugs. Nature New Biology 1971;231(25):232-235. 5. Adams RH. Prostaglandin s, Related Factors, and Cytokines. In: Adams RH, ed. Veterinary Pharmacology and Therapeutics 9th ed. Wiley-Blackwell; 2009:439-455. 6. Boutaud O. Determinants of the cellula r specificity of acetaminophen as an inhibitor of prostagl andin H2 synthases. Proceedings of the National Academy of Sciences 2002;99(10):7130-7135. 7. Schuster VL. Prostaglandin transport. Prostaglandins & Other Lipid Mediators 2002;68-69:633-647. 8. Botting RM. Vane's discovery of the mechanism of action of aspirin changed our understanding of its clinical pharmacology. Pharmacol Rep 2010;62(3):518-525. 9. Smith WL, Langenbach R. Why ther e are two cyclooxygenase isozymes. J. Clin. Invest. 2001;107(12):1491-1495. 10. FitzGerald GA, Loll P. COX in a crystal ball: current status and future promise of prostaglandin research. J. Clin. Invest. 2001;107(11):1335-1337. 11. Papich MG. An Update on Nonsteroidal Anti-Inflammatory Drugs (NSAIDs) in Small Animals. In: Vet Clinics of North Am erica: Small Animal .Vol 38. Elsevier Inc.; 2008:1243-1266. 12. Chandrasekharan NV, Dai H, Roos KL T, et al. COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure, and expression. Proc. Natl. Acad. Sci. U.S.A. 2002;99(21):13926-13931.

PAGE 112

112 13. Lees P, Landoni MF, Giraudel J, Toutain PL. Pharmacodynamics and pharmacokinetics of nonsteroidal anti-in flammatory drugs in species of veterinary interest. J Vet Pharmacol Ther 2004;27(6):479-490. 14. Grosser T, Smyth E, FitzGerald GA Anti-inflammatory, Antipyretic, and Analgesic Agents; Pharmacotherapy of Gout: Introduction. In: Brunton L, Blumenthal DK, Murri N, Hilal-Dandan R, Knollmann BC, eds. The Pharmacological Ba sis of Therapeutics 12nd ed. McGraw-Hill. 15. Kallings P. Nonsteroidal anti-inflammatory drugs. In: Veterinary Clinics of North America: Equine Practice .Vol 9. 3rd ed. 1993:523-541. 16. Bertolini A. Dual acting antiinflammatory drugs: a reappraisal. Pharmacological Research 2001;44(6):437-450. 17. Mukherjee D, Nissen S. Risk of Ca rdiovascular Events Associated With Selective COX-2 Inhibitors. JAMA: the journal of the … 2001. 18. Amer M, Bead VR, Bathon J, Bl umenthal RS, Edwards DN. Use of nonsteroidal anti-inflammatory drugs in patients with cardiovascular disease: a cautionary tale. Cardiol Rev 2010;18(4):204-212. 19. Giorgi M, Cuniberti B, Ye G, et al Oral administration of tepoxalin in the horse: A PK/PD study. Vet J 2010. 20. Cuniberti B, Odore R, Barbero R, et al. In vitro and ex vivo pharmacodynamics of selected non-ster oidal anti-inflammatory drugs in equine whole blood. Vet J 2011. 21. Giorgi M, Mengozzi G, Raffaelli A, Saba A. Characterization of in vivo plasma metabolites of tepoxalin in horses using LC-MS-MS. J Pharm Biomed Anal 2011. 22. Kollias-Baker C, Cox K. Non-steroi dal anti-inflammatory drugs. In: Bertone JJ, Horspool LJI, eds. Equine Clinical Pharmacology 1st ed. London: W.B. Saunders; 2004:247-266. 23. Kahn LH, Styrt BA. Necrotizing so ft tissue infections reported with nonsteroidal antiinflammatory drugs. Ann Pharmacother 1997;31(9):10341039. 24. Maihto TE, Lees P, Taylor JB. Absorption and pharmacokinetics of phenylbutazone in Welsh Mountain Ponies. J Vet Pharmacol Ther 1986;(9):26-39.

PAGE 113

113 25. Lichtenberger LM, Wang ZM, Romero JJ, et al. Non-steroidal antiinflammatory drugs (NSAIDs) associate with zwitterionic phospholipids: insight into the mechanism and reversal of NSAID-induced gastrointestinal injury. Nat. Med. 1995;1(2):154-158. 26. MacAllister CG, Morgan SJ, Borne AT Pollet RA. Comparison of adverse effects of phenylbutazone, flunixin m eglumine, and ketoprofen in horses. J Am Vet Med Assoc 1993;202(1):71-77. 27. Gunson DE, Soma LR. Renal papillary necrosis in horses after phenylbutazone and water deprivation. Vet. Pathol. 1983;20(5):603-610. 28. Kopp KJ, Moore JN, Byars TD. Temp late bleeding time and thromboxane generation in the horse: effects of th ree non-steroidal anti-inflammatory drugs. Equine Vet J 1985;17(4):322-324. 29. Landoni MF, Foot R, Frean S, Lees P. Effects of flunixi n, tolfenamic acid, R(-) and S(+) ketoprofen on the res ponse of equine synoviocytes to lipopolysaccharide stimulation. Equine Vet J 1996;28(6):468-475. 30. Dirikolu L, Woods WE, Boyles J, et al. Nonsteroidal anti-inflammatory agents and musculoskeletal injuries in Thoroughbred racehorses in Kentucky. J Vet Pharmacol Ther 2009;32(3):271-279. 31. Jensen RC, Fischer JH, Cwik MJ. E ffect of age and training status on pharmacokinetics of flunixin meglumine in thoroughbreds. Am J Vet Res 1990;51(4):591-594. 32. Soma LR, Sams R, Duer W, et al. Plasma and serum concentrations of phenylbutazone and oxyphenbutazone in racing Thoroughbreds 24 hours after treatment with va rious dosage regimens. Am J Vet Res 1985;46(4):932-938. 33. Short CR, Sams RA, Soma LR, Tobin T. The regulation of drugs and medicines in horse racing in the Unit ed States. The Association of Racing Commissioners International Uniform Cla ssification of Foreign Substances Guidelines. J Vet Pharmacol Ther 1998;21(2):145-153. 34. Tobin T. Drugs and the Performance Horse 1st ed. Springfield: Charles C Thomas; 1981. 35. Linkous K. Association of Racing Commiss ioners International, Inc 2010:148. 36. Agriculture and Agri-Food Canada. "S chedules of Drugs". Available at: http://www.agr.gc.ca/index_e.php 2006. Accessed May 25, 2011.

PAGE 114

114 37. International Federation of Horser acing Authorities. The International Agreement on Breeding, Racing a nd Wagering. Available at: http://www.ifhaonline.org/ 2009. Accessed May 25, 2011. 38. FEICleanSport.org. "Clean S port" Program. Available at: http://www.feicleansport.org 2010. Accessed May 26, 2011. 39. United States Equestrian Federation. Regulation: Drugs and Medications. Available at: http://www.usef.org 2011. Accessed May 26, 2011. 40. Landoni MF, Lees P. Compar ison of the anti-inflammatory actions of flunixin and ketoprofen in horses appl ying PK/PD modelling. Equine Vet J 4:247256. 41. Gabrielsson J, Weiner D. Pharmacokinetic and Pharmacodynamic Data Analysis: Concepts and Applications 4th ed. Swedish Pharmaceutical Press 42. Scarth JP, Teale P, Kuuranne T. Drug metabolism in the horse: a review. Drug Test Anal 2011;3(1):19-53. 43. AUTHIE EC, GARCIA P, POPOT MA, Toutain PL, DOUCET M. Effect of an endurance-like exercise on the dis position and detection time of phenylbutazone and dexamethasone in the horse: Application to medication control. Equine Vet J 2010;42(3):240-247. 44. Soma LR, Gallis DE, Davis WL, Cochran TA, Woodward CB. Phenylbutazone kinetics and metabolite c oncentrations in the horse after five days of administration. Am J Vet Res 1983;44(11):2104-2109. 45. Landoni MF, Lees P. Pharmaco kinetics and pharmacodynamics of ketoprofen enantiomers in the horse. J Vet Pharmacol Ther 1996;19(6):466-474. 46. Sams R, Gerken DF, Ashcraft SM. Pharmacokinetics of ketoprofen after multiple intravenous doses to mares. J Vet Pharmacol Ther 1995;18(2):108-116. 47. Jaussaud PJ, Besse C, Courtot D, Delatour P. Enantioselective pharmacokinetics of ketoprofen in horses. Journal of Veterinary Pharmacokinetics and Therapeutics (16):373-376. 48. Brideau C, Van Staden C, Chan CC. In vitro effects of cyclooxygenase inhibitors in whole bloo d of horses, dogs, and cats. Am J Vet Res 2001;62(11):1755-1760.

PAGE 115

115 49. Mitchell JA, Akarasereenont P, Thiemermann C, Flower RJ, Vane JR. Selectivity of nonsteroidal antiinfla mmatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc. Natl. Acad. Sci. U.S.A. 1993;90(24):11693-11697. 50. Beretta C, Garavaglia G, Cavalli M. COX-1 and COX-2 inhibition in horse blood by phenylbutazone, fl unixin, carprofen and meloxicam: an in vitro analysis. Pharmacol Res 2005;52(4):302-306. 51. Patrignani P, Panara MR, Greco A, et al. Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J Pharmacol Exp Ther 1994;271(3):1705-1712. 52. Brideau C, Kargman S, Liu S, et al. A human whole blood assay for clinical evaluation of biochemical efficacy of cyclooxygenase inhibitors. Inflammatory Research 1996;45(2):68-74. 53. Laneuville O, Breuer DK, Dewitt DL, et al. Differential inhibition of human prostaglandin endoperoxide H synthas es-1 and -2 by nonsteroidal antiinflammatory drugs. J Pharmacol Exp Ther 1994;271(2):927-934. 54. Rocca B, Secchiero P, Ciabattoni G, et al. Cyclooxygenase-2 expression is induced during human megakaryopoiesis and characterizes newly formed platelets. Proc. Natl. Acad. Sci. U.S.A. 2002;99(11):7634-7639. 55. ARCI. The Association of Racing Commissi oners International's Model Rules of Racing 4th ed. 2010. 56. Kamerling SG, Owens JG. Models fo r assessing the relationship between drug concentration and drug effe ct in performance horses. Br Vet J 1994;150(6):507-525. 57. Toutain PL, Lassourd V. Pharmacokinetic/pharmacodynamic approach to assess irrelevant plasma or urine drug concentrations in postcompetition samples for drug control in the horse. Equine Vet J 2002;34(3):242-249. 58. Tobin T, Harkins JD, Sams RA. Testing for therapeutic medications: analytical/pharmacological relationship s and limitations' on the sensitivity of testing for certain agents. J Vet Pharmacol Ther 1999;22(3):220-233.

PAGE 116

116 BIOGRAPHICAL SKETCH Jennifer Noelle Hatzel was born and raised in Katy, Texas and graduated from High School in the same small town. S he spent three years working towards a Bachelor of Science degree in animal science from Texas Tech University in Lubbock, Texas, but was accepted early to begin ve terinary school in 2003. She successfully obtained a Doctor of Veterinary Medicine with the charter class at Western University of Health Sciences in Pomona, California in May, 2007. Upon graduation, from 2007 to 2008, she completed a rotating hospital in ternship at Peterson and Smith Equine Hospital in Ocala, Florida. In the spring of 2009, she comp leted a specialty fellowship in the area of Neonatal Intensive Care at Hagyard Equine Hospit al in Lexington, Kentucky. Upon completion of the fellowship, she began pursuing the Master of Science degree through the College of Veterinary Medicine at the University of Florida in January, 2010. During her time at the Universi ty of Florida, she became in creasingly interested in the field of equine theriogenology and proceeded by pursuing a residency training program. She obtained and subsequently began this resi dency training in June of 2011, where she currently maintains a position at the Equine Reproductive Laboratory at Colorado State University. She anticipates completing her residency training and sitting boards to become a Diplomate of the Am erican College of Theriogenol ogists in August 2013. She currently resides in Fort Collins, Colo rado with her husband Jeremiah, two dogs (Sedona and Satchel), kitty (S phinx) and longtime equine co mpanion, Winchester.