The main representative of the aniline derivatives is paracetamol (Table 3). This drug possesses only weak antiinﬂammatory but efﬁcient antipyretic ac- tivity. The major advantage of paracetamol lies in its relative lack of serious side effects given that the dose limits are obeyed, although serious events can be observed with low doses, although rarely so (Bridger et al. 1998). A small proportion of paracetamol is metabolized to the highly toxic nucle- ophilic N -acetyl-benzoquinoneimine that is usually inactivated by reaction
a Time to reach maximum plasma concentration after oral administration b Terminal half-life of elimination c Noraminopyrinemethansulfonate-Na, dipyrone d Metabolites of metamizol
with sulfhydryl groups in glutathione. However, following ingestion of large doses of paracetamol, hepatic glutathione is depleted, resulting in covalent binding of N -acetyl-benzoquinoneimine to DNA and structural proteins in parenchymal cells (e.g., in liver and kidney; for review see Seeff et al. 1986). Under these circumstances, dose-dependent, potentially fatal hepatic necrosis may occur. When detected early, overdosage can be antagonized within the ﬁrst 12 h after intake of paracetamol by administration of N -acetylcysteine that regenerates detoxifying mechanisms by replenishing hepatic glutathione stores. Accordingly, paracetamol should not be given to patients with seriously impaired liver function.
Typical indications of paracetamol are fever and pain occurring in the con- text of viral infections. In addition, many patients with recurrent headache
beneﬁt from paracetamol and its low toxicity. Paracetamol is also used in chil- dren, but despite its somewhat lower toxicity in juvenile patients, fatalities due to involuntary overdosage have been reported. According to a recent update of the American College of Rheumatology (ACR) guidelines for osteoarthro-
sis, paracetamol remains ﬁrst-line therapy because of its cost, efﬁcacy, and safety proﬁles (Schnitzer 2002). In fact, paracetamol provides effective relief
of pain for many patients with osteoarthrosis, and has been demonstrated to be safe in a wide range of populations, providing less severe and no gastro- toxic side effects as compared to NSAIDs (Brandt 2003). On the other hand, randomized trials have now demonstrated that NSAIDs and selective COX-2 inhibitors provide superior efﬁcacy compared with paracetamol in patients with osteoarthrosis (Geba et al. 2002; Zhang et al. 2004).
In contrast to NSAIDs and selective COX-2 inhibitors, the analgesic and antipyretic mode of action of paracetamol is still a matter of debate. Evi- dence supporting a central analgesic mode of action has already been demon- strated by the ﬁnding that paracetamol may inhibit nociception-induced spinal prostaglandin synthesis (Muth-Selbach et al. 1999). A few years ago, paraceta- mol was claimed to inhibit a newly discovered COX isoform, derived from the same gene as COX-1 and referred to as COX-3 (Chandrasekharan et al. 2002). However, in this study COX-3 was only detected at relatively small amounts in the cerebral cortex of dogs. Moreover, studies indicating a pharmacologically different regulation of COX-3 as compared to the other two isoenzymes have been performed using canine COX-3 and murine COX-1 and -2 enzymes. In this context it is also noteworthy that COX-3 inhibition by acetaminophen was observed at concentrations above those inhibiting monocyte COX-2 in the hu- man whole blood assay. Meanwhile, the existence of a full-length, catalytically active form of COX-3 in humans has been questioned by two independent groups (Dinchuk et al. 2003; Schwab et al. 2003).
A more plausible mode of action of the drug has been suggested on the basis of observations showing that paracetamol inhibits the COX enzymes by reducing the higher oxidation state of these proteins (Boutaud et al. 2002). In line with this ﬁnding, high levels of peroxides were shown to overcome the inhibitory effect of paracetamol on the COX enzymes in platelets and immune cells. As such, 12-hydroperoxyeicosatetraenoic acid, a major product of platelets, completely reversed the inhibitory action of paracetamol on COX-
1. Likewise, addition of peroxides to interleukin-1-stimulated endothelial cells
(as a model of a peroxide-enriched inﬂammatory tissue) abrogated COX-2 inhibition by paracetamol, whereas paracetamol inhibited COX-2 in these cells in the absence of peroxide at concentrations relevant to its antipyretic effect. Although the peroxide theory needs further conﬁrmation, it may provide a plausible explanation of why paracetamol has only a slight effect on platelet function as well as little or no antiinﬂammatory effect in humans.
Following the discovery of phenazone 120 years ago, the pharmaceutical in- dustry has tried to improve this compound in three ways: Phenazone was chemically modiﬁed to (1) create a more potent compound, (2) yield a water- soluble derivative to be given parenterally, and (3) produce a compound which is eliminated faster and more reliable than phenazone. The results of these at-
tempts are the phenazone derivatives aminophenazone, dipyrone, and propy- phenazone (Table 3). Aminophenazone is not in use anymore because it has been associated with formation of nitrosamines that may increase the risk of stomach cancer. The other two compounds differ from phenazone in their potency and elimination half-life, their water solubility (dipyrone is a water- soluble prodrug of methylaminophenazone), and their general toxicity (propy- phenazone and dipyrone do not lead to the formation of nitrosamines in the acidic environment of the stomach). Phenazone, propyphenazone, and dipy- rone (metamizol) are the predominantly used antipyretic analgesics in many countries (Latin America, many countries in Asia, Eastern and Central Eu- rope). Dipyrone has been accused of causing agranulocytosis. Although there appears to be a statistically signiﬁcant link, the incidence is extremely rare (1 case per million treatment periods; Kaufman et al. 1991; International Agran- ulocytosis and Aplastic Anemia Study 1986; Ibanez et al. 2005). Moreover, all antipyretic analgesics have also been claimed to cause Stevens–Johnson’s syn- drome and Lyell’s syndrome, as well as shock reactions. New data indicate that the incidence of these events is in the same order of magnitude as with, e.g., penicillins (Roujeau et al 1995; Mockenhaupt et al. 1996; Anonymous 1998). All nonacidic phenazone derivatives lack antiinﬂammatory activity, and are devoid of gastrointestinal and (acute) renal toxicity. In contrast to paraceta- mol, dipyrone is safe even when given at overdosages. If one compares aspirin, paracetamol, and propyphenazone when used for the same indication (e.g., occasional headache), it is evident that aspirin is more dangerous than either propyphenazone or paracetamol.
Selective COX-2 Inhibitors
Definition and Structural Basis of COX-2 Selectivity
Selective inhibitors of the COX-2 enzyme, also referred to as coxibs, have been developed as substances with therapeutic actions similar to those of NSAIDs, but without causing gastrointestinal side effects. By deﬁnition, a substance may be regarded as a selective COX-2 inhibitor if it causes no clinically meaningful COX-1 inhibition at maximal therapeutic doses. Among the variety of available test systems, the ex vivo whole blood assay has emerged as the best method to estimate COX-2 selectivity in humans (Patrignani et al. 1997). X-ray crystallog- raphy of the three-dimensional structures of COX-1 and COX-2 has provided more insights into how COX-2 speciﬁcity is achieved. Within the hydrophobic channel of the COX enzyme, a single amino acid difference in position 523 (isoleucine in COX-1, valine in COX-2) has been detected that is critical for the COX-2 selectivity of several drugs. Accordingly, the smaller valine molecule in COX-2 gives access to a side pocket that has been proposed to be the binding site of COX-2 selective substances (Luong et al. 1996).
Comparison of the Pharmacokinetics of Different Selective COX-2 Inhibitors
Selective COX-2 inhibitors currently used are the sulfonamides celecoxib and parecoxib and the methylsulfone etoricoxib. Moreover, the phenylacetic acid derivative lumiracoxib is currently (August 2006) available in ten Latin Amer- ican countries, Australia, New Zealand and Great Britain. Physicochemical and pharmacological data of these compounds as well as those of the COX-2 inhibitors recently withdrawn (rofecoxib and valdecoxib; see Sect. 4.4) are compiled in Table 4.
All selective COX-2 inhibitors are relatively lipophilic compounds. Whereas the sulfonamides and methylsulfones are nonacidic compounds, lumiracoxib is
a phenylacetic acid derivative. Differences in physicochemical characteristics are reﬂected in different pharmacokinetic behavior. Accordingly, the volume of distribution of the nonacidic compounds is equal or above body weight,
whereas that of lumiracoxib is, as with other acetic acid derivatives, around
15% of body weight. The nonacidic compounds distribute almost equally throughout the body with the exception of celecoxib, which is likely to be sequestered in body fat due to its extremely high lipophilicity. The acidic compound lumiracoxib apparently mimics diclofenac and indomethacin that reach high concentrations in the blood stream, kidney, liver, and inﬂamed tissue, but comparatively lower concentrations in other compartments. The physicochemical differences are also reﬂected in other pharmacokinetic pa- rameters. As might be expected from its very high lipophilicity, the absorption of celecoxib is relatively slow and incomplete. The compound undergoes con- siderable ﬁrst pass metabolization (20%–60% oral bioavailability) and its rate
of elimination (t1/2 ∼6–12 h) appears to be highly variable (Werner et al. 2002,
2003). Etoricoxib encounters only slow metabolic elimination (t1/2 ∼20–26 h),
while lumiracoxib is eliminated much faster with an elimination half-life of
about 5 h. Indirect evidence suggests that all substances reach sufﬁciently high concentrations in the central nervous system to counter inﬂammatory pain by blocking COX-2 within about an hour. The fast absorption of etoricoxib appears to cause its fast onset of action. Moreover, all ﬁve compounds are eliminated predominantly by metabolization.
With the exception of rofecoxib, all COX-2 inhibitors undergo oxidative drug metabolism by cytochrome P450 (CYP) enzymes. Interestingly, celecoxib has recently been shown to inhibit the metabolism of the CYP2D6 substrate
metoprolol, the most widely used β-blocker (Werner et al. 2003). This inter- action is likely to occur as well with the concomitant administration of other
CYP2D6 substrates, including sedatives, serotonin reuptake inhibitors, tricyclic antidepressants, and some neuroleptics. On the other hand, the interference of CYP3A4 by etoricoxib appears of minor importance given that CYP3A is
inducible and many drugs metabolized by CYP3A are also substrates of other
Physicochemical and pharmacological data of selective COX-2 inhibitors (data obtained from Brune and Hinz 2004). Despite withdrawal from the market, rofecoxib and valdecoxib are included for comparative reasons
a Ratio of IC50 values (IC50 COX-1/IC50 COX-2) in the human whole blood assay b Volume of distribution c Time to reach maximum plasma concentration after oral administration d Terminal half-life of elimination e All metabolites are less active than the parent compound (except parecoxib) f Compounds may inhibit CYP2D6 g Parecoxib is the water soluble prodrug of valdecoxib releasing valdecoxib within about 15 min, tmax : 30 min h Less at not recommended dosage i Compound may inhibit CYP1A2 k In analogy to diclofenac, the metabolism of lumiracoxib involves probably also other CYP enzymes such as CYP2C8 or CYP2C19 l Lumiracoxib is
currently (August 2006) available in ten Latin American countries, Australia, New Zealand and Great Britain. Intended doses are as follows: 100 mg (symptomatic relief of osteoarthritis); 400 mg (short-term relief of moderate to severe acute pain associated with primary dysmenorrhea, dental surgery , and orthopedic surgery)
On the basis of their diverse pharmacokinetic characteristics, a rational use of different selective COX-2 inhibitors in different clinical settings is recom- mended. Accordingly, the slow absorption and variable ﬁrst pass metaboliza- tion of celecoxib should preclude this compound from the treatment of acute pain. By contrast, etoricoxib appears promising for this indication. However, ﬁnal data are still lacking. This is even more the case with lumiracoxib. Based on experiences gathered with diclofenac, a structural analog, the administra- tion of lumiracoxib in salt form to facilitate fast absorption appears promising for the treatment of acute pain.