Absorption

As lipid-soluble weak organic acids, classical NSAIDs are generally well absorbed when administered orally (Table 20.4), although rate and extent of absorption vary with species, gastric pH, dosing in relation to feeding, and gastrointestinal motility. In monogastric species, absorption from the stomach is favored by the Henderson–Hasselbalch ion trapping mechanism, which maintains a diffusion gradient for undissociated acid molecules between acidic gastric juice and plasma of pH 7.4. In ruminants, initial absorption from the four-compartment stomach and subsequent intestinal absorption creates the basis for a biphasic absorption pattern. In young ruminants, operation of the esophageal groove reflex is another mechanism accounting for double peaks in plasma concentration (Marriner and Bogan, 1979). Despite dilution in a large volume of rumen liquor (approximately 120 liters in an adult cow), bioavailability of phenylbutazone and meclofenamate was shown to be 50–60% in cattle (Lees et al., 1988a; Marriner and Bogan, 1979).

There are exceptions to the high bioavailability of NSAIDs after oral dosing. Bioavailability in horses was <5% for an oil-based formulation of RS-ketoprofen. However, when the pure drug substance was administered, bioavailability was of the order of 50% (Landoni and Lees, 1995). Another factor affecting absorption rate, of significance in the horse, is binding to hay (shown in vitro) and to digesta (demonstrated in vivo) (Lees et al., 1988b). Binding delayed the absorption of phenylbutazone and flunixin, accounting for double peaks in plasma concentration–time curves (Maitho et al., 1986; Welsh et al., 1992). In monogastric species, administration of NSAIDs with food is a common practice, as it may lessen irritant effects on the GI. As in horses, this practice is likely to delay absorption but to reduce absorption only moderately or not at all. However, the bioavailability of mavacoxib was higher in fed (87%) than in fasted (46%) dogs, while the converse was true for robenacoxib in both the dog (62 and 84%, respectively) and cat (10 and 49%, respectively) (Cox et al., 2010; Jung et al., 2009; King et al., 2013). As well as feed, other factors, including product formulation, may influence the rate and/or extent of absorption of NSAIDs. For example, aspirin is available in plain, water-soluble, film-coated, buffered, enteric-coated, and time-release formulations, with differing disintegration and dissolution rates.

Distribution

A potential therapeutic advantage of the high degree of plasma protein binding is accumulation of NSAIDs in inflammatory exudate. Since exudate is rich in albumin, which has leaked from the circulation, the binding of NSAIDs to albumin accounts for ready penetration into and persistence at sites of acute inflammation (Table 20.5). Exudate concentrations of NSAIDs often exceed those in plasma when drug body clearance is high and terminal half-life is short. Accumulation in exudate is a likely explanation for: (i) the maintained effectiveness of NSAIDs when plasma concentrations have decreased to low levels and (ii) why those NSAIDs with short elimination half-lives, for example flunixin, ketoprofen, vedaprofen, tolfenamic acid, and robenacoxib, may be effective with once-daily dosing. This longer duration of effect than predicted from plasma concentration–time profile is associated with the negative hysteresis described in in vivo studies, which indicate that maximal inhibition of PGE₂ synthesis in exudate occurs some time after peak drug concentrations in plasma (Giraudel et al., 2005b; Pelligand et al., 2012, 2014). The latter group reported elimination half-lives from feline exudate for S-ketoprofen and robenacoxib of 26 and 41 hours, respectively; corresponding values for blood were 1.62 and 1.13 hours after subcutaneous administration. Exudate PGE₂ concentration was reduced for 24–36 hours. In clinical OA dogs, the synovial fluid concentration of robenacoxib exceeded the IC₅₀ for COX-2 for 16 hours; the corresponding time for healthy beagles was 10 hours (Silber et al., 2010).

For the 2-arylpropionate NSAIDs, enantioselectivity of distribution into exudate (Table 20.5) and synovial fluid (Armstrong et al., 1999a) has been demonstrated. This impinges on efficacy, as the two enantiomers exhibit pharmacodynamic differences as well (see Section NSAID pharmacodynamics: Inhibition of COX and 5-LO).

For those COXIBs based on sulphonamide or sulphone structures, the molecules are nonionized at physiological pH and volumes of distribution are high, whereas robenacoxib, a carboxylic acid, resembles most nonselective NSAIDs in having a low distribution volume.

A possible consequence of NSAID plasma protein binding is competition with other protein-bound drugs; theoretically this can lead to acute toxicity of either drug as the free concentration is increased. Nephrotic animals have reduced plasma protein concentrations and some authors have suggested that this may predispose to NSAID toxicity. However, in both of these circumstances, this will rarely be clinically significant, as the free concentration quickly falls, being subject to redistribution, metabolism, and excretion (Toutain and Bousquet-Melou, 2002).

The penetration of NSAIDs into milk in the absence of mammary gland infection is poor, milk concentrations being of the order of 1% or less of plasma total concentration. This results from the high degree of binding to plasma protein. Distribution into milk is also limited by the Henderson–Hasselbalch mechanism, as milk pH is less than that of plasma. In mastitis, however, penetration is likely to be increased from these low levels.

Excretion and Metabolism (Elimination)

Examples of elimination half-life are presented in Table 20.6. Interpretation of these values must take account of the fact that half-life is a hybrid variable, controlled by body clearance and volume of distribution.

Species differences in some pharmacokinetic variables are the rule rather than the exception for NSAIDs. For example, while plasma protein binding is almost invariably very high and volumes of distribution are low for classical NSAIDs in most species, clearance and terminal half-life vary markedly. The example of phenylbutazone is presented in Table 20.7. Aspirin, like phenylbutazone, provides a useful example of the impact of elimination on dosing intervals. The terminal half-life of its metabolite salicylate ranges from 32 minutes in cattle to 22–45 hours in the cat, in which species elimination is zero order (dose-dependent pharmacokinetics).

The pharmacokinetics of the 2-arylpropionate subgroup of NSAIDs (carprofen, ketoprofen, and vedaprofen) is influenced by the possession of a single chiral center. There are therefore two mirror-image enantiomeric forms, R and S. The licensed products are the racemic (50 : 50) mixtures. Etodolac, though not a 2-arylpropionate, also possesses a single chiral center. The importance of NSAID chirality derives from the fact that the body is a chiral environment, whose cell membranes, macromolecules, and enzymes are based on D-monosaccharides and L-amino acids. Therefore, while NSAID enantiomers have virtually identical physicochemical properties, they do not have the same pharmacological properties in the body’s chiral environment, as manifested by both pharmacokinetic and pharmacodynamic differences. Pharmacokinetic data based on “the total drug” concept must be viewed with alarm; such data comprises “highly sophisticated scientific nonsense” (Ariens, 1985) because racemic mixtures are simply combinations, in equal proportions, of two distinct drugs. Pharmacokinetic differences are reflected in clearance, elimination half-life, and plasma area under curve (AUC) ratios for the two enantiomers. As illustrated by carprofen, AUC ratios vary between species and may be as high as 4 : 1, despite the 1 : 1 ratio in the administered products (Tables 20.5 and 20.8).

Differences in pharmacokinetics for each enantiomer of chiral NSAIDs may arise in two ways. First, there are enantiomer differences in elimination, arising from differing rates of hepatic metabolism (Soraci et al., 1995). Second, for some drugs chiral inversion occurs in vivo. This, almost invariably, is unidirectional, comprising R to S inversion. Species (and breed) differences in the extent of chiral inversion of ketoprofen have been investigated (Table 20.9). There is also inevitable interanimal variation in the pharmacokinetics of NSAIDs. Table 20.10 presents data for firocoxib in beagles.

As well as interanimal and interbreed differences, a report on celecoxib in a colony of 245 beagle dogs revealed a within-breed difference; two subpopulations were identified, an extensive metabolizer (EM) phenotype for which mean half-life was 1.72 hours and clearance was 18.2 ml/kg/min, and a poor metabolizer (PM) phenotype for which corresponding values were 5.18 h and 7.15 ml/kg/min (Paulson et al., 1999). This example points to the possibility of significant genetic differences in clearance and terminal half-life of other NSAIDs. In fact, similar within-breed (beagle) pharmacokinetic differences (EM and PM phenotypes) have been described for cimicoxib (Jeunesse et al., 2013). Respective clearances were 0.31 and 0.11 l/h/kg. In a lameness model, duration of action was greater in the PM dogs. These authors also administered cimicoxib to dogs of four breeds; half-lives were of the same order of magnitude as those of the EM beagles.

In addition to species, inter- and intrabreed, and interanimal differences in NSAID pharmacokinetics, there are inevitable intraanimal differences, arising from pathological or physiological state. Lascelles et al. (1998) reported, in dogs undergoing anesthesia, greater values of C_max (20.6 versus 11.0 μg/ml) and AUC (175 versus 115 μg/ml·h) when carprofen was administered postoperatively compared to preoperatively. In adult cows with an E. coli endotoxin-induced mastitis, values of AUC and clearance were 507 mg/l·h and 1.4 ml/kg/h for carprofen (Lohuis et al., 1991) whilst values for healthy cows were 294 mg/l·h and 2.4 ml/kg/h. Age may affect the pharmacokinetics of NSAIDs. Clearance is likely to be slower and half-life longer in neonates, as established for phenylbutazone in goats (Eltom et al., 1993) and as suggested for carprofen in young calves (Lees et al., 1996; Delatour et al., 1996) and for phenylbutazone in older horses (Lees et al., 1985). The effect of an endurance exercise on phenylbutazone pharmacokinetics in the horse was investigated by Authie et al. (2010) in relation to medication control; hepatic clearance was decreased, as also was volume of distribution, so that terminal half-life was not changed.

In the dog the clearance of COXIBs differs between the extremes of robenacoxib and mavacoxib (Table 20.11); 13.5 ml/min/kg (robenacoxib) and 0.045 ml/min/kg (mavacoxib) and respective half-lives were 0.63 hours and 17 days in laboratory (beagle or mongrel) dogs. In the cat the elimination half-life of robenacoxib was 0.78 to 1.49 hours and was independent of administration route – intravenous, oral, or subcutaneous (King et al., 2013; Pelligand et al., 2012). Mavacoxib biotransformation and renal excretion in the dog are very limited; elimination is predominantly by biliary secretion and excretion in feces, and enterohepatic recirculation contributes to the slow clearance and long terminal half-life. Lees et al. (2015) calculated that the maximum amount excreted daily in bile was less than 10% of the monthly dose. Based on dose normalized plasma AUC, the pharmacokinetics of mavacoxib administered orally did not differ between mongrel and beagle dogs and was proportional to dose over the range 2–25 mg/kg (Cox et al., 2010).

However, in population pharmacokinetic studies, in mainly elderly large breed dogs with OA, the mean elimination half-life of mavacoxib was longer (44 days) than the half-life of 17 days in laboratory dogs and in approximately 5% of OA dogs T_1/2 exceeded 80 days. This led to reduction in the dose predicted from laboratory dogs of 4 mg/kg to 2 mg/kg, with a dosing interval of 14 days between first and second doses and 28 days between subsequent doses and with a duration of therapy limited to 6 months (Cox et al., 2011). These authors reported body weight as the primary factor predicting clearance and distribution volume (both scaled by bioavailability) and with smaller effects of age and breed. In a clinical population of OA dogs, it was shown that apparent clearance and volume of distribution of robenacoxib were likewise proportional to body weight (Fink et al., 2013). Clearly, there is a need to conduct population pharmacokinetic studies in clinical subjects on other NSAIDs.

Parecoxib is an inactive prodrug, with a short half-life (0.42 hours) after intramuscular injection in the cat; it is converted in vivo to the active metabolite valdecoxib (T_1/2 = 8.52 hours; Kim et al., 2014).

Baert and De Backer (2003) provided valuable comparative pharmacokinetic data for sodium salicylate, flunixin, and meloxicam in five bird species, chicken, ostrich, duck, turkey and pigeon; species differences were marked.

Drug Residues

Inhibition of COX and 5-LO

In addition to pharmacokinetic differences, a second basis for the variation in clinical response to NSAIDs between species, breeds, and individual animals is differences in their pharmacodynamics. This has been studied at several body levels: molecular, cellular, tissue, and whole animal, and can be considered in terms of the action–effect–response relationship (Table 20.12). The principal action is inhibition of COX, the enzyme with a pivotal position in the AA cascade; NSAIDs occupy the hydrophobic channel of COX, preventing AA access to its active site. This leads to the effect of inhibition of synthesis of proinflammatory mediators, including PGE₂ and PGI₂. The action and effect have been demonstrated with clinically achieved concentrations of many NSAIDs in several species (Lees and Higgins, 1985; Landoni and Lees, 1995, 1996; Landoni et al., 1995a,b,c; Cheng et al., 2003; Jones et al., 2002). The responses are the analgesic, antipyretic, antiinflammatory, antithrombotic, and antiendotoxemic properties of NSAIDs (Lees et al., 1986; Welsh and Nolan, 1994; Welsh et al., 1997; Nolan, 2001; Giraudel et al., 2005b). Current evidence suggests that a high level (80% or higher) of inhibition of PGs may be required to achieve good clinical responses.

Vane (1971) discovered that the principal mechanism of action of NSAIDs was inhibition of COX. A further advance occurred with the discovery of two COX isoforms, COX-1 and COX-2 (Kujubu et al., 1991; Xie et al., 1991). It was immediately recognized that most classical NSAIDs inhibit both isoforms, COX-1 inhibition producing toxic effects and COX-2 inhibition providing therapeutic effects. COX-1 was classified as a constitutive enzyme with physiological/ housekeeping functions, including gastro- and renoprotection and blood clotting leading to hemostasis. COX-2 was initially regarded solely as an inducible enzyme, up-regulated at sites of inflammation and responsible for producing proinflammatory mediators. Since 1991, much effort has been directed to identifying the following: which tissues express the two isoforms constitutively; the physiological and pathological roles of each isoform; the role of COX-2 induction at sites of inflammation and centrally; and the nature, incidence, and severity of side effects of drugs that inhibit one or both isoforms.

The overall roles of COX isoforms and the actions of NSAIDs, as perceived in 1991, have been modified as follows:

1. Warner and Mitchell (2003) have concluded that there are only two genes for COX enzymes, COX-1 and COX-2 and not a third, COX-3, as earlier postulated.
   
2. It is now recognized that COX-2 is a constitutive enzyme, present in stomach, brain, spinal cord, kidney, ovary, uterus, and ciliary body. Therefore, it has been postulated that complete inhibition of COX-2, especially over long periods, might be associated with such side effects as abortion, fetal abnormalities, delayed bone healing, delayed healing of soft tissue (including delayed ulcer healing), renotoxicity, and “adverse cardiovascular events.” However, the now extensive clinical use over prolonged periods of COXIBs has been associated with a generally good safety profile. Nevertheless, COXIBs are not free of adverse effects on the GI and there remains controversy over the cardiovascular events associated with their clinical use, especially in human medicine when administered over prolonged periods. COX-2 is both constitutive and inducible in endothelial cells and its selective inhibition might disturb the endothelial PGI₂ (an antiaggregatory and vasodilator PG synthesized via COX-1 and COX-2) platelet TxA₂ (a proaggregatory and vasoconstrictor eicosanoid synthesized by COX-1) balance in the direction of platelet aggregation and vasoconstriction. The cardiovascular events in humans, which have been the subject of much public discussion, might reflect this imbalance. However, some reports have suggested an increased cardiovascular risk in humans receiving long-term therapy with nonselective as well as COX-2 selective drugs. Epidemiological studies in animals reviewing such risks have not been reported.
   
3. Some reports indicate that COX-1 may contribute to the synthesis of proinflammatory PGs. Therefore, as well as COX-2 inhibition, inhibition of COX-1, as provided by the older nonselective NSAIDs, might be required for optimal efficacy. However, this concept is controversial and most experimental and clinical data suggest that selective COX-2 inhibitors are as efficacious as nonselective NSAIDs.
   
4. The introduction of a novel class of NSAIDs, the dual inhibitors, including tepoxalin, which inhibit two enzymes that use AA as a substrate, COX and 5-LO, has failed to make a major impact in veterinary medicine.

Efficacy, Potency, and Sensitivity of Inhibitory Actions of NSAIDs on COX Isoforms

Based on selectivity, NSAIDs are classified as nonselective, COX-1 selective, or as preferential or selective (denoting increasing degrees of selectivity) for COX-2 (Table 20.13). Validating, confirming, and interpreting the data in Table 20.13 is problematic, as the scientific literature reports widely differing potency ratios (COX-1 : COX-2) for individual drugs, even in a single species. The in vitro experimental conditions impact markedly on the COX-1 : COX-2 inhibition ratios, with higher values (i.e., higher selectivity for COX-2) commonly obtained when isolated enzyme, broken cell, or intact cell determinations in buffer are used in comparison with whole blood assays. The latter are accepted as the “gold standard” as they approximate most closely to conditions in the whole animal. Thus, Gierse et al. (2002) using isolated enzyme assays, reported for deracoxib an IC₅₀ COX-1 : COX-2 ratio of 380 : 1 in humans and 1295 : 1 in the dog, whereas McCann et al. (2004) for the same drug, in a canine whole blood assay, obtained a ratio of 12 : 1. Likewise, Ricketts et al. (1998) reported a higher COX-1 : COX-2 ratio for rac-carprofen in isolated cell assays than those obtained in whole blood assays (Table 20.14).

		IC50 ratio COX-	IC80 ratio COX-
Species	Enantiomer	1 : COX-2	1 : COX-2
Human1	RS	0.020       Only gold members can continue reading. Log In or Register to continue Share this: Click to share on Twitter (Opens in new window) Click to share on Facebook (Opens in new window) Related Related posts: Anticonvulsant Drugs Tetracycline Antibiotics Drugs Affecting Animal Behavior Neuromuscular Blocking Agents Stay updated, free articles. Join our Telegram channel Join Tags: Veterinary Pharmacology and Therapeutics Feb 8, 2018 | Posted by admin in PHARMACOLOGY, TOXICOLOGY & THERAPEUTICS | Comments Off on Analgesic, Antiinflammatory, Antipyretic Drugs Full access? Get Clinical Tree Get Clinical Tree app for offline access Get Clinical Tree app for offline access