University of Groningen

Reflections on flurbiprofen eyedropsvan Sorge, Adriaan Alastair

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2002

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):van Sorge, A. A. (2002). Reflections on flurbiprofen eyedrops. Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 10-03-2020

CHAPTER 1GENERAL INTRODUCTION

1.1 FLURBIPROFEN - AN OVERVIEW

IntroductionFlurbiprofen, CAS registry number (Substance name) 5104-49-4, a member of thephenylalkanoic acids (1), a white (or almost white) crystalline powder, melting point114-117°C, practically insoluble in water, but readily soluble in most organic solvents,also known as a hydratropic acid analog (2), was already subjected for evaluation ofits platelet aggregation inhibiting action in 1973 (3). Chemically it is known as: 2-flu-oro-α-methyl-[1,1'-biphenyl]-4-acetic acid; 2-fluoro-α-methyl-4-biphenyl-acetic acid;2-(2-fluoro-4-biphenylyl)propionic acid; 3-fluoro-4-phenylhydratropic acid (4). In1993 its potent anti-platelet activity was evaluated in a double-blind, placebo-con-trolled, multi-centre study for efficacy on preventing reinfarction and reocclusion aftersuccessful thrombolysis or angioplasty in acute myocardial infarction (5).

Flurbiprofen

Flurbiprofen is described in the recent editions of the United States (USP),European (EP), British Pharmacopoeia (BP) and Japanese Pharmacopoeia (JP). Inthe USP both flurbiprofen and its sodium salt are described in the racemic form. Inthe EP the racemic form is also described; however in the EP and BP monographof flurbiprofen the existence of an enantiomer is alluded to ("and enantiomer"). TheJapanese Pharmacopoeia (JPXIII) gives no hint of the chiral nature of the flur-biprofen molecule. A solution of flurbiprofen in methanol giving no optical rotation isthe only description given thereof.

A monograph for flurbiprofen eyedrops is mentioned in the USP as "Flurbiprofensodium ophthalmic solution" and in the BP as "Flurbiprofen eyedrops". They con-tain not less than 90.0% and not more than 110.0% of the prescribed or statedamount. The sodium salt of flurbiprofen in the USP and BP is available in the dihy-drate form.

In the EP, five impurities for flurbiprofen are mentioned. Interestingly 4 are chiral(one chiral centre) and one diastereomeric in nature (two chiral centers).

18

Chapter 1

FCO2H

CH3

PharmacodynamicsThe major pharmacological properties have already been reviewed by Adam et al.in 1975 (1). Several discriminating techniques were applied to determine the lowesteffective oral dose (mg/kg) as anti-inflammatory, analgesic and antipyretic drug. Inthe anti-inflammatory tests three animal species were used: guinea pig, rat andmouse. In the guinea pig the UV-erythema test was employed in which the referencecompound (acetylsalicylic acid, 80 mg/kg) was found to correspond to 0.25 mg/kg offlurbiprofen. In the mouse model the capillary permeability of the peritoneum wasevaluated by use of a dye (Pontamine sky blue). Acetylsalicylic acid at 120 mg/kgwas equivalent to 0.47 mg/kg of flurbiprofen. In the rat three methods wereemployed: the carageenan edema test, and two adjuvant arthritis models for thedeveloping state and the established state. Reference compounds were, respec-tively, acetylsalicylic acid (81 mg/kg) in the first and indomethacin (1 mg/kg) andphenylbutazone (10 mg/kg) in the two latter models. The corresponding lowest effec-tive dose for flurbiprofen was, 0.11 mg/kg, and 0.33 mg/kg in the two latter models.

With the carageenan edema test a subgroup of rats was also tested who werebilaterally adrenalectomised to rule out any adrenocortical interference. Severalconclusions were drawn from this study. Flurbiprofen was devoid of adrenocortical-stimulating properties and was one of the most potent agents of this type reportedyet; at least 10 times more potent than ibuprofen. It was postulated that the modeof action in the mouse and rat was not identical to that of acetylsalicylic acidIn US patent 3,755,427 (August 28th 1973) (6) it was stated that flurbiprofen wasbetween 75 to over 100 times as potent as acetylsalicylic acid.

In (2) the relative potency of various hydratropic acids were tested for their relaxingability on guinea pig tracheal ring contraction after sensitization by rat SRS-A.Furthermore the paper not only provided information for flurbiprofen but also for thelevorotary (-) and dextrorotary (+) enantiomers. It became apparent that the relaxingpotency of the racemic mixure (±) was unexpectedly too low as compared to the dex-trorotary component suggesting that the dextrorotary component was hindered by thesimultaneous presence of the levorotary component. The putative interactionbetween the two enantiomers was tested by the simultaneous addition of the two sep-arate enantiomers to the muscle bath. Reversal by the dextrorotary component wasdiminished by the simultaneous presence of the (-) flurbiprofen. Taking this in consid-eration (+) flurbiprofen was approximately 80 fold more potent than (-) flurbiprofen.

PharmacokineticsPharmacokinetic properties have been assessed in different species (7). In man(8), when assessed by HPLC of the racemic molecule, a two-compartment openmodel appeared the most appropriate for flurbiprofen. Drug absorption efficiencywas found independent of the oral dose. The intact drug resides mainly in the

19

General introduction

peripheral and central compartments, disappearing with a terminal half life ofapproximately 5.5 hours. More than 99% of flurbiprofen is bound to serum proteins.The serum flurbiprofen concentrations in clinical use however show an occupancyof less than 10% of the primary binding sites. The binding site differs from that ofdrugs like oral anticoagulants and sulphonamides. Drug interactions will thereforenot automatically occur with simultaneous use.

Oxidation and conjugation are the main pathways of metabolism. More than 95%of an oral dose is excreted via the kidney within 24 hours. Forty to 47% of a dailyoral dose is excreted as 2-[2-fluoro-4'-hydroxy-4-biphenylyl]propionic acid; 5% as 2-[2-fluoro-3',4'-hydroxy-4-biphenylyl]propionic acid; 20-30% as 2-[2-fluoro-3'-hydroxy-4'-methoxy-4-biphenylyl]propionic acid and 20-25%. as the parent mole-cule flurbiprofen. Between 65 - 85% of flurbiprofen and its metabolites are presentas glucuronide and sulfate conjugates.

Stereoselective HPLC of human plasma has also been performed (9). After oraladministration of 25 mg of the R(-) enantiomer of flurbiprofen no indication wasfound that inversion to the S(+) enantiomer occurred. This was confirmed in healthyvolunteers taking either 50 mg R(-) flurbiprofen or S(+) flurbiprofen (10). Severalstudies on the pharmacokinetics of flurbiprofen in the rat have been performed allshowing that in this species a minimal amount of inversion could take place(approx. 5%), the inversion halftime being approximately half an hour (11,12,13,14).

Stereoselectieve studies have been performed following the disposition of flur-biprofen in normal volunteers after a single 50 mg racemic dose (15), in healthyfemale subjects following oral administration of the single enantiomers of flurbipro-fen, 50 mg S(+) flurbiprofen or R(-) flurbiprofen or 100 mg R(-) flurbiprofen or place-bo, in a 4-way crossover design with placebo (16); in patients with end-stage renaldisease undergoing continuous ambulatory peritoneal dialysis (CAPD) after admin-istration of a single 100 mg racemic dose (17), and stereoselective disposition ofracemic flurbiprofen in single and multiple dosing in uraemic patients (18). On thebasis of pharmacokinetics, adjustment of flurbiprofen dosing in uraemic patients isnot necessary. In CAPD patients circulating plasma levels of flurbiprofen proved 40-50% lower than in normal subjects implying that analgesia could be less thanexpected in this selected group of patients. Accumulation of the 2-[2-fluoro-4'-hydroxy-4-biphenylyl]propionic acid metabolite, which has minimal anti-inflammato-ry activity, does occur in this group of patients but the clinical significance is notestablished. In patients with liver disease with ascites and in renal failure patientswith a creatinine clearance of less than 10 ml.min-1, significant higher free fractionsof R(-)- and S(+) flurbiprofen were detected in conjunction with lower albumin con-centrations (19). An overview of the clinical pharmacokinetics of flurbiprofen and itsenantiomers is presented in (20).

20

Chapter 1

21

A different model was introduced for the investigation of the pharmacokinetics offlurbiprofen enantiomers and the simultaneous inhibition of prostanoid production(21). This study was performed in healthy volunteers in whom, after receiving oral-ly either 75 mg R(-), S(+) flurbiprofen or no medication in a randomised 3-waycross-over design, flurbiprofen pharmcokinetics were analysed by HPLC andprostanoid production was monitored by enzyme immuno assay and chemilumi-nescence assay. Here also no clinically relevant inversion of R(-) to S(+) flurbipro-fen was seen. However, the study showed unexplained discrepancies in severalstages of the pharmacokinetic and pharmacodynamic parameters of the flurbipro-fen enantiomers.

Chiral inversion of R(-) flurbiprofen to S(+) flurbiprofen has been studied in vitroto investigate the mechanism behind this phenomenon which is not shared by all 2-arylpropionic acids (22). With crude rat liver homogenates it was demonstrated thatan acyl-CoA synthase enzyme in conjunction with ATP and Mg2+ is obligatory. Thefirst step comprises the metabolic formation of a CoA thioester of R(-) ibuprofen. Ithas been made plausible that this step takes place in adipose tissue since afteradministration of R(-) flurbiprofen significant amounts of both enantiomers arefound in adipose tissue. By way of a non-stereoselective racemase (epimerase) thisproduct is converted to its S(+) ibuprofen CoA thioester. Through action of a hydro-lase, S(+) ibuprofen is released from its CoA thioester form. This unidirectionalenantioselective chiral inversion in man has not been reported for flurbiprofen,carprofen and ketoprofen (23,24,25). Research on the enzymatic inversion at thechiral carbon atom had been done earlier by use of deuterated ibuprofen. In thatstudy it was noted that the R(-) isomer is the only substrate for the epimerisationreaction (26). A summary of the metabolic chiral inversion of 2-arylpropionic acidderivatives, the variations between species and the complexity that can arise dueto formed chiral metabolites has been presented in (27).

Biliary excretion has been examined in normal and bile-duct cannulated rats forthe enantiomers of flurbiprofen after intravenous dosing of 10 mg/kg of each enan-tiomer (28). It was found that the fraction of enterohepatic circulation was greaterfor R(-) flurbiprofen than for its antipode. Although the S(+) flurbiprofen enantiomerwas excreted to a greater extent in bile, reabsorption from the intestine was insignif-icant. One reason for this phenomenon may be the presumed stereoselectivehydrolysis of the flurbiprofen conjugates with preference for the R(-) enantiomer.However in later similar experiments (29) enterohepatic cycling of both flurbiprofenenantiomers could be demonstrated.

Glucuronidation in rat and human liver microsomes proceeds faster for the R(-)enantiomer than for its S(+) antipode. Glucuronidation is facilitated by the enzymecomplex UDP-glucuronosyltransferase of which there exist several isoforms (30).However not the identity of the isoenzyme but the stereoselective interaction of theenantiomer influences the reaction velocity.

General introduction

In a review (31) on the binding of flurbiprofen to albumin in human plasma it wasreported that at low therapeutic concentrations the S(+) enantiomer has a higherprotein binding than its R(-) antipode. At high drug concentrations there is no meas-urable difference, however. In an ultrafiltration study done with normal volunteersthe free fraction of R(-) flurbiprofen was higher than its S(+) antipode at low druglevels but similar for both enantiomers at higher drug levels. Patients with renalimpairment and patients exhibiting hypoalbuminaemia have higher free fractions offlurbiprofen enantiomers than normal volunteers. Plasma protein binding of anenantiomer is not influenced by its own concentration or the presence of itsantipode under clinical therapeutic conditions (32).

In a model study using isolated perfused rabbit lungs it was demonstrated thatflurbiprofen does not undergo pulmonary metabolism to any extent (33).

As mentioned above (8) the main routes of biotransformation of flurbiprofen arethrough oxidation and conjugation. Oxidation has been investigated more specifi-cally (34) for the enantiomers of flurbiprofen utilizing human liver microsomes. Themost prominent oxidative metabolism route is by cytochrome P450. It was estab-lished that cytochrome P450 2C9 and its allelic variant R144C catalysed the oxida-tive reaction. Interestingly, there was no stereoselective preference of one enan-tiomer over the other.

Safety for intestinal permeability changes when using the racemate or the sepa-rate enantiomers of flurbiprofen was studied in rats for which species it was estab-lished that only a minimal inversion of the R(-) enantiomer takes place. Intestinalpermeability was measured by urinary excretion of 51Cr-EDTA (35). It was estab-lished that at both dosages used (1 mg/kg and 3 mg/kg for the racemic drug andhalf for the enantiomers) permeability was significantly different from control. R(-)flurbiprofen was safest in both dosage ranges. S(+) flurbiprofen inflicted similardamage as the racemic form.

In (36) it was shown that in rats R(-) flurbiprofen gave the same increase of intes-tinal permeability, but the difference was that the impact on mucosal prostanoid pro-duction was smaller and not accompanied by ulcerative changes in the small intes-tine.

Although it would seem attractive to develop therapeutic R(-) enantiomers of 2-arylproionic acids due to its supposedly lower toxicological profile it must be bornein mind that the presumed pharmacological action required for reducing inflamma-tion is inhibition of prostaglandin synthesis. This property resides primarily, in thecase of flurbiprofen, in the S(+) enantiomer for which a difference of 30 to 100 timescompared to the R(-) enantiomer was established depending on the model used.Only with full metabolic inversion of a R(-) enantiomer to a S(+) enantiomer wouldsuch a therapeutic drug be a possibility. For flurbiprofen this is not the case inhumans (37,38,39).

22

Chapter 1

In a comparative study (40) done in rabbits, the inhibitory effect on rise in intraocu-lar pressure and increase in aqueous humor protein after topical application ofarachidonic acid (5% in peanut oil) by 14 nonsteroidal anti-inflammatory inhibitorswas measured. For 50% inhibition of the intraocular pressure response, flurbipro-fen ranked second best with an effective concentration of approximately 0.06%.Indomethacin (suspension in water, not further specified) ranked 4th with an approx-imate concentration of 0.2%.

In a short review (41) the importance of the involvement of prostaglandins to cer-tain eye conditions is discussed. The rise in intraocular pressure and the break-down of the blood-aqueous barrier were related to these compounds. A search forthe best drug in inhibiting prostaglandin mediated diseases was called for beforetesting them in the human eye.

The comparative in vivo inhibitory effects of flurbiprofen, indomethacin and acetyl-salicylic acid, all as sodium salt solutions, have been tested in the rabbit anterioruvea and conjunctiva after topical (0.5% solutions) and intraperitoneal administra-tion (42). In both methods of administration acetylsalicylic acid almost completelyabolished prostaglandin synthesis. Flurbiprofen given intraperitoneally was morepotent than indomethacin which inhibited prostaglandin synthesis only partially,even at twice the dose of acetylsalicylic acid and flurbiprofen. Topical administrationrevealed that acetylsalicylic acid performed well even at a dose as low as 0.01%but indomethacin and flurbiprofen performed poor.

Use of flurbiprofen (0.01% and 0.1%) was evaluated in comparison to 1% pred-nisolone as an inhibitor of corneal neovascularization in New Zealand albino rab-bits (43). Flurbiprofen 0.1% and prednisolon 1% were equally effective in inhibitingvessel growth.

As an alternative possibility for the use of topical administration of corticosteroidsa nonsteroidal anti-inflammatory drug was considered (44). Flurbiprofen was test-ed in a double-blind fashion to see if intraocular pressure would change and if itsuse could block corticosteroid induced ocular hypertension. In a selected group ofpatients, with known intraocular sensitivity towards corticosteroids, flurbiprofen eye-drops (0.03%) did not alter intraocular pressure following six weeks of treatment.Also pretreatment by flurbiprofen did not block corticosteroid-induced ocular hyper-tension.

Flurbiprofen was also investigated for human use in the prevention of intraocularinflammation (45). In a randomised double-blind parallel group study, placebo orflurbiprofen (100 mg thrice daily) was given orally for 8 days starting 24 hoursbefore routine cataract extraction. Flurbiprofen was only favoured over placebo forthe resolution of corneal inflammation at day 6. Interestingly, flurbiprofen concen-trations in the aqueous were detected up to 4 hours after the last dose with a con-

23

General introduction

centration of 0.57 mg/L (2.3x10-6M). The authors conclude that flurbiprofen may beof value in the treatment of uveitis and other kinds of intraocular inflammation. In a study (46) involving female New Zealand rabbits, eyedrop disposition was stud-ied after application of 14C labeled flurbiprofen in a concentration of 0.03%. Nometabolism was detected for flurbiprofen in the eye. The total amount present inocular tissues (cornea, aqueous humor, iris, ciliary body, choroid and retina) in nor-mal rabbit eyes 30 minutes after application of 50 microliter of a 0.03% solution,was 4.25%. At 6 hours this was 1.59%. Following ocular application 77.51±8.79%was found in a 24-urine collection period. Unchanged flurbiprofen accounted for25.3±3.6%.

Ocular availability was studied in female albino rabbits (47) receiving 50 microliterof 0.3% or 0.15% solution of flurbiprofen. The ocular bioavailability of the 0.3% solu-tion was 10% and for the 0.15% solution 7%. The elimination half-life in the aque-ous humor was 93 minutes which approximates the turnover rate of aqueous humorin the rabbit and indicates that drainage is the main route of elimination.

Bioavailability was determined after a single dose and after multiple doses oflabeled flurbiprofen in rabbit eyes using topical application (48). Multiple dosing ofa flurbiprofen solution of 0.03%, every half hour three doses, gave levels in the eyehigh enough to prevent prostaglandin synthesis. It compared favourably to the useof a single drop of 0.1% solution possibly because of less irritation of the eye andthus less stimulation of tear flow. Peak tissue concentrations were reached between30 minutes and 60 minutes and were 2 to 6 times higher in all tissues than seenafter one drop of 0.1% solution.

To determine the intraocular concentration of flurbiprofen sodium in the humanaqueous humor of patients undergoing cataract surgery, samples were taken afterreceiving flurbiprofen sodium at selected times prior to surgery (49). Only singledrop instillation was done. Samples of aqueous humor were analysed by HPLC.Flurbiprofen concentrations were detectable in the aqueous between 30 minutes to7.25 hours after topical application.

To determine if and how much drug can penetrate to the posterior segment of theeye, a study was done in white New Zealand albino rabbits, by single drop methodof dosing, using paracetamol 1% in saline and bendazac lysine 0.5% in saline orother solvent (50). When comparing the data with literature data it appeared thatparacetamol behaved similar to flurbiprofen as regards penetration into the aque-ous humor, having a very poor entry into the vitreous and attaining higher concen-trations in the retina than paracetamol. In all three eye compartments bendazaclysine permeated poorly. The data suggest an alternative entry route to the poste-rior segments of the eye. It appears that the lens acts as a barrier for the entry fromthe aqueous.

24

Chapter 1

In a study done by Carabaza et al. inhibition of prostaglandin synthesis was inves-tigated using the enantiomers of three NSAIDs (ketoprofen, flurbiprofen and ketoro-lac), including stereoselective inhibition of inducible COX-2 (51). It became appar-ent that inhibition by the three enantiomer pairs is comparable for COX-1 and COX-2. With both cyclooxygenase isoenzymes inhibition resides almost exclusively inthe S(+) isomer.

One of the most frequent problems encountered during cataract surgery is invol-untary pupillary constriction. In the past, several pharmacological interventionshave been tried as a remedy, but without success. Albino rabbits have been usedto study the effect of topical administration of indomethacin (1% aqueous solutionwith no further specification of buffer and pH used) and flurbiprofen (0.03% aque-ous solution of sodium flurbiprofen) on this unwanted condition (52). In this set-upalso local anaesthetics, capsaicin, sympathomimetic agents and an anticholinergicwere involved all according to a specified protocol. Flurbiprofen demonstrated asignificant inhibitory effect on miosis while topical indomethacin failed. However nosingle agent or combination of agents blocked the miotic response completely.

Although nonsteroidal anti-inflammatory drugs are pharmacologically effectiveinhibitors of cyclooxygenase activity and prostaglandin synthesis (53), cyclooxyge-nase-independent anti-inflammatory actions of NSAIDs are also known (54). Sinceit was reported that sodium salicylate and acetylsalicylic acid inhibit the action of thetranscription factor nuclear factor kappa B (NF-κB), the enantiomers of flurbiprofenwere tested in a zymosan-induced paw inflammation model. Although R(-) flur-biprofen does not inhibit cyclooxygenase to a significant extent, it is more potentthan S(+) flurbiprofen and almost as effective as dexamethasone in this inflamma-tory model. Inhibition of NF-κB by R(-) flurbiprofen resulted in a reduced expressionof COX-2 and tumor necrosis factor a (TNF-α).

Nitric oxide formed by the inducible NO synthase (iNOS) has been implicated asa mediator of pain and tissue injury in various inflammatory and autoimmune dis-eases. In an in vitro model involving RAW 264.7 macrophages, it could be demon-strated that iNOS mRNA expression is equipotently suppressed by the enantiomersof flurbiprofen. S(+) flurbiprofen and R(-) flurbiprofen did not inhibit LPS inducedCOX-2 mRNA expression but did inhibit LPS-induced prostaglandin E2 formationenantioselectively, with the S(+) antipode being 46 times more active than the R(-)flurbiprofen (IC50 0.0061µM and 0.28 µM respectively). Collectively, these findingswould suggest that the pharmacological (i.e. anti-inflammatory) activity of the flur-biprofen enantiomers is not only related to inhibition of cyclooxygenase enzymeactivities but also to inhibition of transcription factor activition like NF-κB and AP-1,resulting in diminished formation of pro-inflammatory factors like iNOS and TNF-α(55,56).

25

General introduction

26

REFERENCES

1. Adam SS, McCullough KF, Nicholson JS. Some biological properties of flurbiprofen, an anti-inflammatory, analgesic and antipyretic agent. Arzneim Forsch (Drug Research) 1975;25:1786-91.

2. Greig ME, Griffin RL. Antagonism of slow reacting substance in anaphylaxis (SRS-A) and otherspasmogens on the guinea pig tracheal chain by hydratropic acids and their effects on anaphy-laxis. J Med Chem 1975;18:112-116.

3. Nishisawa EE, Wynalda DJ, Suydam DE Molony BA. Flurbiprofen, a new potent inhibitor ofplatelet aggregation. Thrombosis research 1973;3:577-88.

4. The Merck Index, Thirteenth Edition, 2001, Merck & Co., Inc. Whitehouse Station, N.J., USA. 5. Brochier ML. Evaluation of flurbiprofen for prevention of reinfarction and reocclusion after suc-

cessful thrombolysis or angioplasty in acute myocardial infarction. The flurbiprofen French trial.Eur Heart J 1993;14:951-7.

6. US patent 3,755,427 (August 28th, 1973).7. Risdall PC, Adams SS, Crampton EL, Marchant B. The disposition and metabolism of flurbiprofen

in several species including man. Xenobiotica 1978;8:691-704.8. Kaiser DG, Brooks CD, Lomen PL. Pharmacokinetics of flurbiprofen. Am J Med 1986;80:10-5

(suppl 3A).9. Jamali F, Berry BW, Tehrani MR, Russell AS. Stereoselective pharmacokinetics of flurbiprofen in

humans and rats. J Pharm Sci 1988;77:666-9.10. Geisslinger G, Menzel-Soglowek S. Stereoselective high-performance liquid chromatographic

determination of flurbiprofen in human plasma. J Chromatogr 1992;573:163-7.11. Knihinicki RD, Day RO, Graham GG, Williams KM. Stereoselective disposition of ibuprofen and

flurbiprofen in rats. Chirality 1990;2:134-40.12. Jamali F, Berry BW, Wright MR. Dose-dependency of flurbiprofen enantiomer pharmacokinetics

in the rat. J Pharm Sci 1994;83:1077-80.13. Peskar BM, Kluge S, Peskar BA, Soglowek SM, Brune K. Effects of pure enantiomers of flur-

biprofen in comparison to racemic flurbiprofen on eicosanoid release from various rat organs exvivo. Prostaglandins 1991;42:515-31.

14. Berry BW, Jamali F. Enantiomeric interaction of flurbiprofen in the rat. J Pharm Sci 1989;78:632-4.15. Knadler MP, Brater DC, Hall SD. Stereoselective disposition of flurbiprofen in normal volunteers.

Br J Clin Pharmacol 1992;33:369-75.16. Geisslinger G. Stereoselective disposition of flurbiprofen in healthy subjects following adminis-

tration of the single enantiomers. Br J Clin Pharmacol 1994; 37:392-4.17. Cefali EA, Poynor WJ, Sica D, Cox S. Pharmacokinetic comparison of flurbiprofen in end-stage

renal disease subjects and subjects with normal renal function. J Clin Pharmacol 1991;31:808-14.18. Knadler MP, Brater DC, Hall SD. Stereoselective disposition of flurbiprofen in uraemic patients.

Br J Clin Pharmacol 1992;33:377-83.19. Blouin R, Chaudhary I, Nishikara K, Cox S. The effects of liver and renal disease on stereoselective

serum binding of flurbiprofen. Br J Clin Pharmacol 1993;35:62-4.20. Davies NM. Clinical pharmacokinetics of flurbiprofen and its enantiomers. Clin Pharmacokinet

1995;28:100-14.21. Oelkers R, Neupert W, Williams KM, Brune K, Geisslinger G. Disposition and effects of flurbipro-

fen enantiomers in human serum and blister fluid. Br J Clin Pharmacol 1997;43:145-53.22. Knihinicki RD, Williams KM, Day RO. Chiral inversion of 2-arylpropionic acid non-steroidal anti-

inflammatory drugs-1. In vitro studies of ibuprofen and flurbiprofen. Biochem Pharmacol1989;38:4389-95.

23. Mayer JM. Stereoselective metabolism of anti-inflammatory 2-arylpropionates. Acta Pharm Nord1990;2:197-216.

Chapter 1

27

24. Knadler MP, Hall SD. Stereoselective arylpropionyl-CoA thioester formation in vitro. Chirality1990;2:67-73.

25. Wechter WJ. Drug chirality: on the mechanism of R-aryl propionic acid class NSAIDs.Epimerization in humans and the clinical implications for the use of racemates. J Clin Pharmacol.1994;34:1036-42.

26. Wechter WJ, Loughhead DG, Reischer RJ, VanGiessen GJ, Kaiser DG. Enzymatic inversion at sat-urated carbon: nature and mechanism of the inversion of R(-) p-iso-butyl hydratropic acid. BiochemBiophys Res Commun. 1974;61:833-7.

27. Caldwell J, Hutt AJ, Fournel-Gigleux S. The metabolic chiral inversion and dispositional enan-tioselectivity of the 2-arylpropionic acids and their biological consequences. Biochem Pharmacol1988;105-14.

28. Menzel S, Beck WS, Brune K, Geisslinger G. Stereoselectivity of biliary excretion of 2-arylpropi-onates in rats. Chirality 1993;5:422-7.

29. Eeckhoudt SL, Evrard PA, Verbeeck RK. Biliary excretion and enterohepatic cycling of R- and S-flurbiprofen in the rat. Drug Metab Dispos 1997;25:428-30.

30. Hamdoune M, Mounie J, Magdalou J, Masmoudi T, Goudonnet H, Escousse A. Characterizationof the in vitro glucuronidation of flurbiprofen enantiomers. Drug Metab Dispos 1995;23:34308.

31. Lapicque F, Muller N, Payan E, Dubois N, Netter P. Protein binding and stereoselectivity of non-steroidal anti-inflammatory drugs. Clin Pharmacokinet 1993; 25:115-25.

32. Knadler MP, Brater DC, Hall SD. Plasma protein binding of flurbiprofen: enantioselectivity and influ-ence of pathophysiological status. J Pharmacol Exp Ther 1989; 249:378-85.

33. Hall SD, Hassazadeh-Khayyat M, Knadler MP, Mayer PR. Pulmonary inversion of 2-arylpropionicacids: influence of protein binding. Chirality 1992;4:349-52.

34. Tracy TS, Rosenbluth BW, Wrighton SA, Gonzalez FJ, Korzekwa KR. Role of cytochrome P4502C9 and an allelic variant in the 4’-hydroxylation of (R)- and (S) flurbiprofen. Biochem Pharmacol1995;49:1269-75.

35. Davies NM, Wright MR, Russell AS, Jamali F. Effect of the enantiomers of flurbiprofen, ibuprofen,and ketoprofen on intestinal permeability. J Pharm Sci 1996;85:1170-3.

36. Mahmud T, Somasundaram S, Sigthorsson G, Simpson RJ, Rafi S, Foster R, Tavares IA, RosethA, Hutt AJ, Jacob M, Pacy J, Scott DL, Wrigglesworth JM, Bjarnason I. Enantiomers of flurbipro-fen can distinguish key pathophysiological steps of NSAID enteropathy in the rat. Gut1998;43:775-82.

37. Wright MR, Davies NM, Jamali F. Rationale for the development of stereochemically pure enan-tiomers: are the R enantiomers of chiral nonsteroidal anti-inflammatory drugs inactive? J PharmSci 1994;83:911-2.

38. Menzel-Soglowek S, Geisslinger G, Beck WS, Brune K. Variability of inversion of (R)-Flurbiprofenin different species. J Pharm Sci 1992;81:888-91.

39. Geisslinger G, Menzel-Soglowek S, Beck WS, Brune K. R-flurbiprofen: isomeric ballast or activeentity of the racemic compound? Agents Actions Suppl 1993;44:31-6.

40. Podos SM, Becker B. Comparison of ocular prostaglandin synthesis inhibitors. Invest Ophthalmol1976;15:841-4.

41. Podos SM. Prostaglandins, nonsteroidal anti-inflammatory agents and eye disease. Trans AmOphthalmol Soc 1976;74:637-60.

42. Kulkarni PS, Srinivasan D. Comparative in vivo inhibitory effects of nonsteroidal anti-inflammato-ry agents on prostaglandin synthesis in rabbit ocular tissues. Arch Ophthalmol 1985;103:103-6.

43. Cooper CA, Bergamini MVW, Leopold IH. Use of flurbiprofen to inhibit corneal neovasculariza-tion. Arch Ophthalmol 1980;1102-5.

44. Gieser DK, Hodapp E, Goldberg I, Kass MA, Becker B. Flurbiprofen and intraocular pressure.Ann Ophthalmol 1981;13:831-3.

General introduction

28

45. Hillman JS, Frank GJ, Kheskani MB. Flurbiprofen and human intraocular inflammation. InAdvances in prostaglandin and thromboxane research. Vol 8;1723-5:1980. Edited by B.Samuelsson, P.W. Ramwell, and R. Paoletti. Raven Press New York, USA.

46. Anderson JA, Chen CC, Vita JB, Shackleton M. Disposition of topical flurbiprofen in normal andaphakic rabbit eyes. Arch Ophthalmol 1982;100:642-5.

47. Tang-Liu DD-S, Liu SS, Weinkam RJ. Ocular and systemic bioavailability of ophthalmic flur-biprofen. J Pharmacokinet Biopharm 1984;12:611-26.

48. Anderson JA, Chen CC. Multiple dosing increases the ocular bioavailability of topically adminis-tered flurbiprofen. Arch Ophthalmol 1988;106:1107-9.

49. Ellis PP, Pfoff DS, Bloedow DC, Riegel M. Intraocular diclofenac and flurbiprofen concentrationsin human aqueous humor following topical application. J Ocular Pharmacol 1994;10:677-82.

50. Romanelli L, Morrone LA, Guglielmotti A, Piccinelli D, Valeri P. Distribution of topically adminis-tered drugs to the posterior segment of rabbit eye. Pharmacol Res 1992;25: 39-40 (Suppl 1).

51. Carabaza A, Cabre F, Rotlan E, Gomez M, Gutierrez M, Garcia ML, Mauleon D. Stereoselectiveinhibition of inducible cyclooxygenase by chiral nonsteroidal antiinflammatory drugs. J ClinPharmacol 1996;36:505-12.

52. Duffin RM, Camras CB, Gardner SK, Pettit TH. Inhibitors of surgically induced miosis.Ophthalmology 1982;89:966-79.

53. Versteeg HH, van Bergen en Henegouwen PMP, van Deventer SJH, Peppelenbosch MP.Cyclooxygenase-dependent signalling: molecular events and consequences. FEBS 1999;445:1-5.

54. Tegeder I, Pfeilschifter J, Geisslinger G. Cyclooxygenase-independent actions of cyclooxygenaseinhibitors. FASEB J 2001;15:2057-72.

55. Tegeder I, Niederberger E, Israr E, Gühring H, Brune K, Euchenhofer C, Grösch S, Geisslinger G.Inhibition of NF-κB and AP-1 activation by R- and S-flurbiprofen. FASEB J 2001;15:2-4; and 595-7. To read full text: http://www.fasebj.org/ cgi/doi/10.1096/fj.00-0130fje.

56. Hinz B, Brune K, Rau T, Pahl A. Flurbiprofen enantiomers inhibit inducible nitric oxide synthaseexpression in RAW 264.7 macrophages. Pharm Res 2001;18:151-6.

Chapter 1

1.2 CATARACT AND CATARACTOGENESIS

CataractAlterations in lens transparency increase with age. It seems possible that the lensstays transparent until the age of 120 years. In the fifth decade of life howeverapproximately 65% of people will have some form of lens opacity. This can varyfrom small spots to complete opacification. The patient will not immediately noticethe development of cataractogenesis as this process does not proceed with anovert inflammatory process nor is any pain experienced. Symptoms accompanyingsuch a process are difficulty in reading, in recognizing faces, watching television,seeing in bright light and during driving (1,2). A simple test is available to assessvisual function, the Snellen chart, but reliability warrants consideration (3).

On a global scale cataract is the commonest cause of visual disability and by farthe single largest cause of blindness (4). Traditional eye medicines in rural Africainflict corneal ulcers and cause blindness in children in a quarter of cases (5). Thebest known medication to cause cataracts are corticosteroids (6) with evidence thatphenothiazines, amiodarone, chloroquine and possibly acetylsalicylic acid alsomight be associated with increased risk (7). There are still limitations in the identifi-cation of the causes of cataracts, not only in developing countries but also in indus-trialised countries (8). Research is ongoing to gain a better understanding of thegenetics of human cataract. It can be envisaged that knowledge of congenitalcataract will provide more insight into the putative role of genes in age-relatedcataract (9).

There is still no effective, pharmacological, remedy for established cataractsalthough a theoretical and experimental basis is building up to address age-onsetcataractogenesis by anti-cataract agents (10).

Treatment is purely surgical with an established success rate >90%. Two basictechniques are in use for management: extracapsular cataract surgery and intra-capsular cataract surgery.

The surgical removal of an opacified lens was first reported in 1745 (April 8th,Marseille) performed by Jacques Daviel (1693-1762). In 1752 two lectures werepresented by him at the Académie Royale de Chirurgie in Paris where an accountwas given of 206 lens extractions. Of these 182 were successful: 88%. Almost twocenturies later Sir Harold Ridley performed the first successful lens implantation(London, November 29th 1949).

Innovation is still improving cataract surgery, especially by the technologicaladvancement of extracapsular extraction and posterior-chamber intraocular lensimplantation (11). These substantial improvements should become available to anincreasing group of patients on the waiting list (12,13,14). Outpatient cataract sur-

29

General introduction

gery seems very well possible without loss of quality (15). Bilateral cataract extrac-tion can be safely done within 48 hours (16).

CaractogenesisAge related, or senile cataract is the most common form and inflicts blindnessworldwide. There are two types of age-related cataract, nuclear and cortical. Oneof the possibilities that has been investigated for nuclear caractogenesis is throughhydroxyl radical-attack of lens proteins which causes cross-linking and proteinaggregation, ultimately resulting in opacity of the lens (17). In another model it isproposed that cataract is essentially a conformational disease in which non-enzy-matic modification of amino groups e.g. by sugars and steroids destabilize the lensproteins and causes conformational changes. The interaction between the aminoacid of a lens protein and a sugar, well known as the Maillard reaction, will not onlygive rise to a colored reaction product but will also cause the protein to cross-link,aggregate and eventually to become insoluble which in turn will opacify the lens(18). In another sugar-related process it was investigated whether the polyol path-way was involved in the process of cataractogenesis. By the enzyme aldose reduc-tase glucose can be converted to sorbitol. However, when a limited amount ofantioxidants is available a significant amount of hydrogen peroxide can be formed.This will give rise to the production of hydroxyl radicals and will lead to the initialstage of a "sugar" cataract (19). Another reported mechanism on the formation ofcataract was the kynurenine metabolic pathway. The tryptophan metabolite, 3-hydroxykynurenine is present at elevated concentrations in the lens and is able toabsorb UV radiation. However, an excessive amount in the lens has been report-edly associated with cataract formation. In the rabbit eye enzymes, leading to theformation of this metabolite, are present in the iris/ciliary body. The formed metabo-lite is taken up by the lens for formation of UV-filtering products. If however anexcess of 3-hydroxykynurenine is present in the lens, free radical formation mayoccur, which will ultimately lead to tissue injury like lens opacification (20,21).Oxidative damage of lens proteins seems to be a major factor in cataract formation.A threshold of protein oxidation has been identified at which opacification will takeplace (22). Radiation-induced cataractogenesis will commence above 1 Gy as hasbeen observed in survivors of the Hiroshima and Nagasaki atomic bombs (23).

The lens is not a purely passive optical element but maintains an internal circula-tion for lens transparency. Sodium-potassium pumps have been identified in thelens as well as a major intrinsic protein belonging to the aquaporin family of waterchannels. Glucose is transported from the aqueous humor to the lens for energysupport. It has been postulated that dysfunction of any of these links in this chainof events may ultimately lead to cataract formation (24).

30

Chapter 1

31

General introduction

REFERENCES

1. Elkington AR, Khaw PT. Cataracts. BMJ 1988;296:1787-90.2. Cotlier E. The Lens in: Adler's Physiology of the eye. Clinical application. Chapter 10; 6th Edition

1975. RA Moses, MD Editor. The CV Mosby Company ISBN 0-8016-3540-3.3. McGraw P, Winn B, Whitaker D. Reliability of the Snellen chart (Better charts are now available).

BMJ 1995;1481-2.4. Thylefors B, Négrel A-D, Pararajasegaram R, Dadzie KY. Global data on blindness. Bull WHO

1995;73:115-21.5. Anonymous. Traditional eye medicines: a note of concern. WHO Drug information 1995;9:152-3.6. Butcher JM, Austin M, Mc Galliard J, Bourke RD. Bilateral cataracts and glaucoma induced by

long term use of steroid eyedrops. BMJ 1994;309:43.7. Cumming RG, Mitchell P. Medications and cataract (The blue mountains eye study).

Ophthalmology 1998;105:1751-8.8. Johnson GJ. Limitations of epidemiology in understanding pathogenesis of cataracts. Lancet

1998;351:925-6,9. Francis PJ, Berry V, More AT, Bhattacharya S. Lens biology; development and human catarac-

togenesis. TIG 1999;15:1916.10. Benedek GB, Pande J, Thurston GM, Clark JI. Theoretical and experimental basis for the inhibi-

tion of cataract. Progress in retinal and eye research. 1999;18:391-402.11. Tielsch JM. Appropriate technology for cataract surgery. Lancet 1998;352:754-5.12. Gray CS, Crabtree HL, Oçonnell JE, Allen ED. Waiting in the dark: cataract surgery in older peo-

ple (We need a better means of assessing priorities for surgery). BMJ 1999;318:1367-8.13. Allan B. Intraocular lens implants (Have come a long way, but the advances are not yet available

to all). BMJ 2000;320:73-4.14. Fielder AR, Watson MP, Seward HC, Murray PI. Action on cataracts should influence surgical

training. BMJ 2000;321:639.15. Javitt JC, Street DA, Tielsch JM, Wang Q, Kolb MM, Schein O, SommerA, Bergner M, Steinberg

EP, on behalf of the Cataract Patient Outcomes Research team. Ophthalmology 1994;101:100-6.16. Booth A, Coombes A, Rostron C. Bilateral cataract extraction can be safely done within 48 hours.

BMJ 1999;319:579.17. Fu S, Dean R, Southan M, Truscott R. The hydroxyl radical in lens nuclear catractogenesis. J Biol

Chem 1998;273:28603-9.18. Crabbe MJC. Cataract as a conformational disease - the maillard reaction, alpha-crystallin and

chemotherapy. Cell Mol Biol 1998;44:1047-50.19. Kubo E, Miyoshi N, Fukuda M, Akagi Y. Cataract formation through the polyol pathway is asso-

ciated with free radical production. Exp Eye Res 1999;68:457-64.20. Chiarugi A, Rapizzi E, Moroni F, Moroni F. The kynurenine metabolic pathway in the eye: studies

on 3-hydroxykynurenine, a putative cataractogenic compound. FEBS1999;453:197-200.21. Davies MJ, Truscott RJW. Photo-oxidation of proteins and its role in cataractogenesis. J Photo

Biol 2001;63:114-25.22. Boscia F, Grattagliano I, Vendemiale G, Micelle-Ferrari T, Altomar E. Protein oxidation and lens

opacity in humans. Invest Ophthalmol Vis Sci 2000;41:2461-5.23. Belkacémi Y, Touboul E, Méric JB, Rat P, Warnet JM. Cataract radio-induit: aspects phys-

iopathologiques, radiobiologiques et cliniques. Cancer/Radiother 2001;5:397-412.24. Donaldson P, Kistler J, Mathias RT. Molecular solutions to mammalian lens transparency. News

Physiol Sci 2001;16:118-23.

1.3 CYSTOID MACULAR EDEMA

In 1942 a report was published on a toxic ocular reaction. The characteristic findingwas that the primary aqueous (aqueous humor obtained on a first paracentesis) didcoagulate in contrast to the usual finding that coagulation only takes place in sec-ondary aqueous (aqueous humor obtained on a second paracentesis). The natureof this phenomenon in the - as it was termed - plasmoidtoxic aqueous- of the primaryaqueous, was investigated in some detail. Although no exact culprit could be definedit became clear that the increased permeability of the ciliary body was a major con-tributing factor in the toxic ocular reaction and that the fibrinogen system played anessential role in the coagulation process of the plasmoidtoxic aqueous (1).

In 1953 a paper was published in which a complication was described followingcommon intracapsular cataract extraction. One of the features of this complicationwas the development of postoperative macular changes, and of ultimate reductionof vision as a result of vitreous opacities or macular degeneration. Of the 1,068cataract extractions 894 were intracapsular extractions. Of these 483 occurredintact; the remaining 222 showed complications varying from marked prolapse of thevitreous into the anterior chamber without rupture to late rupture of the anteriorhyaloid with or without adhesions. The percentage of patients encountered with poorvision as a result of vitreous opacities or macular degeneration was found to be 2%.This was similar to the postoperative detachment of the retina after cataract surgery,as found in a total of reviewed 1,200 cases (2).

In 1966 a new study was presented showing the advantage of the use of intra-venous sodium fluoresceinate to detect the lesion. It was demonstrated that resolu-tion of fluorescein leakage into the retina and optic nerve generally parallels the clin-ical resolution of edema of the macula and optic disc. The earlier reports of an inci-dence of 2% of cystoid macular edema was questioned and it was expected to behigher as experience was gained with this new staining technique (3).

A review on the complication of cystoid macular edema in 1976 showed that thiscomplication following cataract surgery was the most common and troublesome (4).The progress in surgical techniques was impressive enough to diminish the majori-ty of complications other than cystoid maculopathy. The incidence of clinically sig-nificant cystoid macular edema remained 2 - 6%. In an attempt to grasp the etiolog-ical factors the author (A.R. Irvine) put forward the possibility of "vasoactive factorsfrom inflammatory cells in the vitreous to penetrate the retina preferentially at themacula and disc". Of the possibilities to produce aphakic cystoid macular edema,inflammation and increased permeability were major steps in the reaction sequence.Medication seemed straight forward as to use steroids. However oral therapy provedof transient value just like periocular steroid injections and had unfavourable sideeffects. Topical steroid therapy was found to be ineffective. New perspectives

32

Chapter 1

appeared following the elucidation of the role of prostaglandins in inflammatory vas-cular permeability changes. However the controlled study mentioned in this reviewusing indomethacin (orally 25 mg tid for 3 weeks) failed to demonstrate any benefi-cial effect (4).

In 1985 a hypothesis was put forward for aphakic cystoid macular edema. Basedon the results of a randomised double blind trial that showed a reduction in incidenceof 50% for aphakic cystoid macular edema by use of an ultraviolet radiation-absorb-ing chromophore in a posterior chamber intraocular lens, it was postulated that post-operative exposure to near-ultraviolet radiation generates free radicals. These radi-cals would facilitate the synthesis of inflammatory mediators like prostaglandins.Prostaglandins are involved in the breakdown of blood-ocular barriers. It follows thata combination of factors like UV-A radiation and the synthesis of prostaglandins is apossibility worth testing as a contributing factor toward cystoid macular edema andtherefore amenable to medical treatment (5).

In an update of the pharmacological therapy it was mentioned that topical non-steroidal anti-inflammatory agents were still not commercially available. However,topical indomethacin was mentioned as the one agent effective in the prophylaxis ofangiographic aphakic cystoid macular edema. Other nonsteroidal anti-inflammatoryagents and corticosteroids are mentioned but no evidence was presented other thananecdotal, not detailed enough or in number too small to evaluate statistically (6).

Further studies on cystoid macular edema revealed that any disturbance of the vit-reous can lead to this syndrome. In particular three possibilities are mentioned bywhich intraocular lenses can give rise to chronic cystoid macular edema (7). Theseare iris chaffing in combination with an uveitis-glaucoma-hyphema syndrome afterposterior chamber intraocular lens implantation, movement of the intraocular lens withintermittent corneal touchings and the corneo-retinal inflammatory syndrome com-promising both the cornea and the retina. When one of these three situations occurintraocular lens removal is required to prevent permanent macular damage (7).

Cystoid macular edema has also been described in the French literature as Irvine-Gass syndrome. An extensive treatise is presented in (8).

In the German literature a report was published on the safety and efficacy of a 1%indomethacin suspension for the prevention of cystoid macular edema (also knownas Irvine-Gass-Norton syndrome or Irvine syndrome). The incidence of cystoid mac-ular edema was 1.34%. Side effects of the eyedrops, as observed in 10% of thecases, were mainly conjunctival in origin (9). Another pharmacological approach forfailing visual acuity, local application of steroids and an injection of tolazoline (α-adrenergic antagonist, having some cholinergic, H2-histaminergic, and 5HT1 recep-tor antagonistic properties as well) in Tenon's capsule, is proposed (10).

In the meantime the FDA has approved several topical NSAIDs for clinical use inophthalmology (11). The approvals are restricted to specific indications, however;

33

General introduction

flurbiprofen sodium and suprofen for the prophylaxis of surgical miosis, ketorolac forthe relief of itching due to allergic conjunctivitis and diclofenac for the treatment ofpostcataract inflammation. For intraoperative miosis no conclusive evidence hasbeen presented that an NSAID is effective. For the prevention of postcataract surgi-cal inflammation the NSAIDs are at least as effective and perhaps more effectivethan corticosteroids in preventing disruption of the blood-aqueous barrier. For cys-toid macular edema the evidence is that topical NSAIDs are better than topical cor-ticosteroids.

In a Canadian report the incidence of aphakic/pseudophakic cystoid macularedema in 90 studies from 1979 to 1991 is presented using three different techniques(12). For intracapsular and extracapsular cataract extraction and the phacoemulsifi-cation technique it varied between 2 - 10%, 0 - 7.6% and 0.6 - 6.0%, respectively.However when using fluorescein angiography the incidence varied between 40 -60%, 2.7 - 11.3% and 2.1 - 6.0%, respectively. Interestingly, aphakic cystoid macu-lar edema occurs more frequently with intracapsular than extracapsular cataractextraction, and even less with placement of an intraocular lens in an intact capsularbag. It seems that the capsular bag encompasses properties other than just for sup-port. The lens barrier protects, it seems, against access of inflammatory agents intothe vitreous. Permanent visual impairment due to cystoid macular edema will varybetween 0.5% and 2%. Treatment of clinically or angiographically proven cystoidmacular edema with indomethacin decreased the incidence of cystoid macularedema; however there was no difference in visual outcome between active andplacebo treated groups. No long-term effectiveness was shown yet with treatment bya NSAID. In the same report carbonic anhydrase inhibitors (e.g. acetazolamide) arementioned as possibly effective drugs in cystoid macular edema caused by changesin the external blood-retinal barrier (retinitis pigmentosa). However, a recent reporton gastric mucosa samples obtained by biopsy showed that NSAIDs (acetylsalicylicacid, indomethacin, naproxen, diclofenac and piroxicam) can activate the carbonicanhydrase isoenzymes I, II and IV (13).

An extensive review was published in 1998 (14) in which the view is held that theincidence, pathogenesis and treatment of cystoid macular edema following cataractsurgery are still poorly understood. Incidence of cystoid macular edema is greatestfollowing an intracapsular cataract extraction with implantation of an iris clip lens inan older population with systemic vascular disease. Clinical characteristics of cys-toid macular edema are a nonuniform distribution of the retinal intravascular fluidwithin the macula leading to accumulation of transudate and ultimately to a sympto-matic or asymptomatic decrease in visual acuity. Preferential leakage from perifovealcapillaries in eyes with cystoid macular edema cannot be explained yet and possi-bly reflects a result of an unknown capillary vitreous interaction. Inflammation, how-ever, is the mainstay in the development of cystoid macular edema. Presumably,

34

Chapter 1

35

breakdown of the blood aqueous barrier is associated with the development of cys-toid macular edema. This was established in rabbits in which the topical activity ofNSAIDs in stabilizing the blood aqueous barrier following paracentesis was studiedby monitoring the integrity of the barrier using anterior ocular fluorophotometrybefore and after paracentesis. The integrity was followed by changes in fluoresceinconcentrations measured after intravenous administration of fluorescein sodium inthe anterior chamber of the eye. All NSAIDs studied (flurbiprofen 0.03%, 0.1%diclofenac, 0.5% ketorolac and 1% suprofen) stabilized the blood aqueous barrierafter paracentesis.

In a Swiss overview it was concluded that NSAIDs with potent anti-inflammatoryproperties allow good control of ocular inflammation, effective maintenance of mydri-asis during surgery and delay the onset of cystoid macular edema (15). A recentreview on topical NSAIDs for ophthalmic use concluded that the benefit-risk ratio isstill favorable when they are applied in an appropriate and judicious manner (16).

REFERENCES

1. Ayo C. A toxic ocular reaction. II On the nature of the reaction. J Immunol 1942;46:127-32.2. Irvine SR. A newly defined vitreous syndrome following cataract surgery. Am J Ophthalmol

1953;36:599-619.3. Gass JDM, Norton EWD. Cystoid macular edema and pailledema following cataract extraction.

Arch Opthalmol 1966;76:646-61.4. Irvine AR. Cystoid maculopathy. Surv Ophthalmol 1976;21:1-17.5. Jampol LM. Aphakic cystoid macular edema; a hypothesis. Arch Ophthalmol 1985;103:1134-5.6. Jampol LM. Pharmacologic therapy of aphakic and pseudophakic cystoid macular edema.

Ophthalmology 1985;92:807-10.7. Drews RC. The present understanding of cystoid macular oedema. Trans opthalmol Soc UK

1985;104:744-7.8. Sole P et al (Expertise bibliographique). Le syndrome d' Irvine Gass. J Fr Ophthalmol 1986;1:75-83.9. Dirscherl M, Straub W. Zur prophylaxe des zystoiden Makulaödems nach Katarkatoperationen

(eine anwendungsbeobachtung van Chibro-Amuno 3). Ophthalmologica 1990;200:142-9.10. Hruby K. Das Irvine-syndrom; diagnose, pathogenese und therapie. Fortschr Ophthalmol

1985;82:147-8. 11. Jampol LM, Jain S, Pudzisz B, Weinreb RN. Nonsteroidal anti-inflammatory drugs and cataract

surgery. Arch ophthalmol 1994;112:891-4.12. Rocha G, Deschenes J. Pathophysiology and treatment of cystoid macular edema. Can J

Ophthalmol 1996;31:282-8.13. Puscas C, Chis F, Pasca R, Pasca S, Mihaescu M, Puscas I. Gastric, vascular and antidiuretic

effects of indomethacin are dependent on direct activation of carbonic anhydrase (CA). Gut1999;45:S5:P0087.

14. Flach AJ. The incidence, pathogenesis and treatment of cystoid macular edema followingcataract surgery. Tr Am Ophth Soc 1998;96:557-634.

15. Guex-Crosier Y. Anti-inflammatoires non stéroïdiens (AINS) et inflammation oculaire. KlinMonatsbl Augenheilkd 2001;218:305-8.

16. Gaynes BI, Fiscella R. Topical nonsteroidal anti-inflammatory drugs for ophthalmic use. A safetyreview. Drug Saf 2002;25:233-50.

General introduction

1.4 PROSTANOIDS

In 1929 it was documented that feeding of rats not only should include essentialelements like amino acids, vitamins and minerals but also small amounts of unsat-urated fat. Analysis uncovered the essential fatty acid linoleic acid, an eighteen car-bon atom chain with two double bonds.

Screening of many types of polyunsaturated fatty acids showed arachidonic acid(20 carbon atoms and 4 double bonds) to be the most active fatty acid to preventmanifestations of nutritional deficiency.

In 1930 a factor was discovered in human semen that contracted the uterus; italso lowered blood pressure. Von Euler demonstrated this to be a fatty acid andintroduced in 1935 the name prostaglandin. In 1960 Bergström elucidated the struc-ture of some prostaglandins, one of them being prostaglandin E2. In 1964 van Dorpcarried out experiments in the Unilever Research laboratories (Vlaardingen) anddemonstrated by use of labeled arachidonic acid that it was converted intoprostaglandin E2 by the medium of homogenized sheep seminal vessels. This find-ing was published in 1964 jointly with Bergström in the same journal. In 1970 Vanepublished his discovery that the biosynthesis of prostaglandins was inhibited byaspirin-like drugs. Four years later Samuelsson discovered the bioconversion byplatelets of arachidonic acid into thromboxane A2. Two years later prostacyclin wasdiscovered by Vane.

The first comprehensive reviews were published in 1974 summarizing resultsfrom the already expanding field of prostaglandin research (1,2). All efforts werenow geared to 'visualize' the prostaglandin endoperoxide synthase enzyme(Cyclooxygenase, COX). Known features were that cyclooxygenase is a polypep-tide, homodimeric in nature (approximately 70 kDa) and in monotopic arrangementin the cell membrane. It carried two distinct functional enzyme activities, catalyzingboth the bisoxygenation of arachidonic acid to its hydroperoxy arachidonatemetabolite prostaglandin G2 and consecutively catalyzing the peroxidative reduc-tion of prostaglandin G2 to its endoperoxide H2 (3). The peroxidase activity of theenzyme complex is not affected by NSAIDs.

A major advance in the field of eicosanoid research was the discovery of a sec-ond inducible cyclooxygenase isoenzyme, COX-2 (4,5). The two isoforms, COX-1and COX-2, were believed to explain the therapeutic but also the adverse effects ofthe frequently used NSAIDs. By hypothesizing that the COX-1 enzyme was theconstitutive enzyme, designed to be available for physiological functions, the COX-2 enzyme was thought primarily to act in pathophysiological processes (COXdogma). An important aim would be to develop COX-2 specific NSAIDs that wouldaid in fighting inflammatory processes and not displaying unwanted side-effectsrelated to inhibition of the COX-1 enzyme (6,7).

36

Chapter 1

Evidence was delivered that localization of the COX-1 enzyme primarily was in thesmooth endoplasmic reticulum and the COX-2 enzyme in the nuclear envelope.Cell membrane receptors for prostanoids have been localized on several tissuesincluding the eye (8,9).

In studying the dynamics of inhibition by flurbiprofen and indomethacin of thehuman prostaglandin H synthases it became evident that these compounds influ-enced at least five processes, including the rate of catalytic activation, the rate ofsubstrate turnover, the rate of autoinactivation of the enzyme complex and theassociation and dissociation rates of the inhibitor with the complex. Overall,indomethacin and flurbiprofen behaved similarly towards the human prostaglandinendoperoxide H synthase-1 and -2 enzymes, although the individual kinetic param-eters differed (10,11). In an earlier study it was concluded that the inhibitor-enzymecomplex is more stable for the flurbiprofen-prostaglandin H synthase-1 than for flur-biprofen-prostaglandin H synthase-2 complex (12).

Each isoenzyme, prostaglandin endoperoxide H synthase-1 and prostaglandinendoperoxide H synthase-2, is encoded by a different gene. When activated andwith adequate arachidonic acid and oxygen present, a single prostaglandinendoperoxide H synthase (COX) molecule can produce 103 molecules ofprostaglandin G2, a hydroperoxide, which is catalytically reduced to its alcoholicform (PGH2) by peroxidase (13). The COX-1 and COX-2 enzymes are homodimers.Each dimer consists of three independent folding units: a membrane-bindingdomain, an enzymatic/catalytic domain and an epidermal growth factor-like domain.The enzymatic/catalytic domain consists of two separate, closely spaced, but inter-dependent areas encompassing 80% of the protein. The cyclooxygenase site is sit-uated at the apex of a long hydrophobic channel in the prostaglandin endoperoxideH synthase molecule. A marked difference between COX-2 and COX-1 is the larg-er channel and the approximately 20% larger binding site in the former. This differ-ence has been exploited to examine the possibility of designing selective COX-2inhibitors (14). Although it has been suggested that the COX-1 enzyme complex ismainly active for maintenance purposes and the COX-2 enzyme, by nature of itsrapid induction capabilities, for the contribution to prostaglandin related inflamma-tion, pain and fever, implying that selective COX-2 inhibitors would benefit thepatient, some doubt has arisen concerning this elegant theory (15,16). In a morerecent study using a COX-1 selective inhibitor as well as COX-2 selective and non-selective inhibitors in normal and monoarthritic rats and mice with paw inflamma-tion, it was concluded that inhibition of both COX-isoenzymes was needed for effec-tive analgesia in inflammation (17). In studies involving the use of cyclooxygenaseknockout mice it became apparent that deficiency of COX-2 had more pronouncedeffects on the physiological maintenance of the body than deficiency of COX-1 (18).To initiate the cyclooxygenase reaction, activation of the peroxidase active site is

37

General introduction

necessary (19). When activated, an aqueous insoluble, nonchiral, arachidonic acidmolecule, liberated by phospholipase A2 from the membrane phospholipids, will be'sucked' into the hydrophobic channel where it will be converted to theprostaglandin G2 endoperoxide. After bioconversion of arachidonic acid into theprostaglandin G2 endoperoxide it will be transported to a reservoir type enclaveformed by the dimeric cyclooxygenase enzym complex. From here transport ofprostaglandin G2 to the peroxidase catalytic site is possible, where it will be trans-formed into prostaglandin H2 endoperoxide.

Closer examination of the interaction of the NSAID flurbiprofen, a representativeof the 2-phenylpropionic acid class, with the enzyme channel, reveals that theamino acid tyrosine 355, situated near the entrance of the channel, creates a localnarrowing which in turn provides a handsome explanation why the S(+) flurbiprofenenantiomer of this molecule will have a better fit than its counterpart R(-) flurbipro-fen (20,21). The carboxylate group of flurbiprofen as well as of indomethacin willcomplex with the guanidinium group of arginine 120 just like the carboxylate groupof arachidonic acid. Preparation of neutral NSAIDs by transforming the carboxylategroup into an ester or amide function may enhance COX-2 selectivity (22).

The cyclooxygenase enzymes produce prostanoids from the polyunsaturated fattyacid, arachidonic acid. These prostaglandins, D2, E2, F2α, I2 and thromboxane A2, actvia their respective receptors to elicit various physiological reactions (23).

A number of receptors have been identified (24). Prostaglandins are involved,together with histamine and bradykinine, in the local increase in vascular perme-ability and edema in which PGE2 and PGI2 are prominently involved. It has becomeevident that they elicit these vascular changes as well as inflammatory pain via EP-and IP-type receptors, respectively.

In a study involving knock-out mice with a EP1 receptor deficient status, the pain-sensitivity responses were tested in two acute prostaglandin-dependent models (25).The animals' reaction was reduced by approximately 50%, the same amount ofreduction that could be achieved by pharmacological interventions in wild-type mice.

In determining the inhibitory effects of the flurbiprofen enantiomers on the COX-1and COX-2 isoenzymes, it is of advantage to take the biological surroundings of theNSAID in the human body, e.g. causing protein binding, into consideration. Use ofhuman whole blood, as an ex vivo method for quantification of the inhibitory effectsof an NSAID on the synthesis of prostanoids has become an established procedure(26,27). As an unequivocal example of functional COX-1 the Ca2+-ionophore stimu-lated platelet is used, measuring the metabolite of thromboxane A2. Thelipopolysaccharide stimulated monocyte can be taken as an activity indicator offunctional COX-2. Any interference by platelets is excluded by acetylation of theCOX-1 enzyme by prior administration of acetylsalicylic acid.

38

Chapter 1

Prostaglandin levels have been measured using radioimmunoassays in 41 humaneyes of patients undergoing vitrectomy (28). In 'quiet eyes' undergoing routinecataract extraction physiological prostaglandin levels of around 100 pg/ml werereported. In patients with a diagnosis of cataract and cystoid macular edema meanlevels of PGI2 varied between 49 and 360 pg/ml.

Induction of COX-2 mRNA can take place in the rabbit eye within three hoursfollowing glaucoma filtration surgery (29). Paracentesis however fails to induceCOX-2 mRNA, possibly because of the minimal disturbance by this procedure.

Prostaglandin synthase activity exists both in vascular and avascular structures ofthe eye, being most abundant in the iris-ciliary body.

Although the iris-ciliary body of the rabbit eye has a high functional capacity tosynthesize prostanoids following paracentesis, only a transient ocular inflammato-ry response follows which resolves within 3 - 4 hours (30).

Eicosanoid measurements in the aqueous obtained during paracentesis of therabbit eye showed a strong and rapid rise in PGE2 levels in the aqueous humor withpeak values at 20 minutes, followed by recovery to baseline within 48 hours (31).No prostacyclin (measured as the stable metabolite 6-keto-PGF1α) was detected inthe aqueous humor at the start but was markedly present at ten and twenty minutesafter the initial trauma. Prostaglandin synthesis was followed shortly by an increasein aqueous humor protein, with peak levels achieved within 30 minutes after para-centesis. Both PGE2 and protein levels declined gradually to near baseline levels48 hours after trauma.

Experimental evidence has shown that the inducible COX-2 mRNA is present in thefirst hours after injury and is possibly assisting in wound healing (32,33). COX-1knockout mice do not show signs of spontaneous gastrointestinal ulceration aswould have been expected when the generation of prostaglandins by the COX-1enzyme is involved in maintenance and integrity keeping purposes (34). ClassicalNSAIDs still show efficacy when the COX-2 enzyme is no longer present, as shownby Western blotting (35). Administering COX-2 inhibitors then does not seem indi-cated. Interestingly, a COX-2 enzyme complex showed up again near resolution ofinflammation and produced anti-inflammatory prostaglandins, PGD2, PGF2α and amember of the PGJ2 family (36). This is in contrast with the view hinting thatcyclooxygenase-2 inhibitors might be the new approach to therapy in ocular inflam-mation (37).

The case against the use of COX-2 inhibitors seems more compelling now that ithas become clear that in the field of rheumatology a five fold higher risk of cardiaccomplications can occur (38,39,40,41). In a well designed study using knockout

39

General introduction

mice deficient of a prostacyclin receptor or a thromboxane receptor, it was shownthat efficient cross talk exists between prostacyclin- and thromboxane- dependentsignaling pathways (42). Inhibition of one signaling pathway might induce the other.This could account for unforeseen complications with the use of cyclooxygenase-2inhibitors (43). Also the presence of mutations in cytochrome P450 isoenzymes thatmetabolize arachidonic acid could play a role in diseases involving clotting andinflammatory disorders (44). Potential drug alternatives, however, are being devel-oped (45).

Flurbiprofen is an NSAID with preferential cyclooxygenase-1 inhibiting capacitylike indomethacin. Its use as a NSAID eyedrop for combatting inflammatoryresponses following cataract surgery therefore seems appropriate.

REFERENCES

1. Flower RJ. Drugs which inhibit prostaglandin biosynthesis. Pharmacol Rev 1974;26:33-67.2. Flower RJ, Vane JR. Inhibition of prostaglandin biosynthesis. Biochem Pharmacol 1974;23:1439-50.3. Merlie JP, Fagan D, Mudd J, Needleman P. Isolation and characterization of the complementary

DNA for sheep seminal vesicle prostaglandin endoperoxide synthase (cyclooxygenase). J BiolChem 1988;263:3550-3.

4. Xie W, Chipman JG, Robertson DL, Erikson RL, Simmons DL. Expression of a mitoogen-respon-sive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad SciUSA 1991;88:2692-6.

5. Xie W, Robertson DL, Simmons DL. Mitogen-inducible prostaglandin G/H synthase: a new targetfor nonsteroidal anti-inflammatory drugs. Drug Dev Res 1992;25:249-65.

6. Vane J. Towards a better aspirin. Nature 1994;367:215-6.7. Allen KN. Aspirin - now we can see it. Nature Medicine 1995;1:882-3.8. Goetzl EJ, An S, Smith WL. Specificity of expression and effects of eicosanoid mediators in nor-

mal physiology and human diseases. FASEB J 1995;9:1051-8.9. Morita I, Schindler M, Regier MK, Otto JC, Hori T, DeWitt DL, Smith WL. Different intracellular

locations for prostaglandin endoperoxide H synthase-1 and -2. J Biol Chem 1995;270:10902-8.10. Callan OH, So O-Y, Swinney DC. The kinetic factors that determine the affinity and selectivity for

slow binding of human prostaglandin H synthase 1 and 2 by indomethacin and flurbiprofen. J BiolChem 1996;271:3548-54.

11. So O-Y, Scarafia LE, Mak AY, Callan OH, Swinney DC. The dynamics of prostaglandin H syn-thases. Studies with prostaglandin H synthase 2 Y355F unmask mechanisms of time-dependentinhibition and allosteric activation. J Biol Chem 1998;273:5801-7.

12. Laneuville O, Breuer DK, Dewitt DL, Hla T, Funk CD, Smith WL. Differential inhibition of humanprostaglandin endoperoxide H synthases-1 and -2 by nonsteroidal anti-inflammatory drugs. JPharmacol Exp Ther 1994;271:927-34.

13. Kulmacz RJ. Cellular regulation of prostaglandin H synthase catalysis. FEBS letters1998;430:154-7.

14. Brooks P, Emery P, Evans JF, Fenner H, Hawkey CJ, Patrono C, Smolen J, Breedveld F, Day R,Dougados M, Ehrich EW, Gijon-Baños J, Kvien TK, van Rijswijk MH, Warner T, Zeidler H.Interpreting the clinical significance of the differential inhibition of cyclooxygenase-1 andcyclooxygenase-2. Rheumatology 1999;38:779-88.

15. Wallace JL. Selective COX-2 inhibitors: is the water becoming muddy? TiPS 1999;20:4-6.

40

Chapter 1

16. Reuter BK, Asfaha S, Buret A, Sharkey KA, Wallace JL. Exacerbation of inflammation-associat-ed colonic injury in rat through inhibition of cyclooxygenase-2. J Clin Invest 1996;98:2076-85.

17. Mazario J, Gaitan G, Herrero JF. Cyclooxygenase-1 vs. cyclooxygenase-2 inhibitors in the induc-tion of antinociception in rodent withdrawal reflexes. Neuropharmacology 2001;40:937-46.

18. Langenbach R, Loftin C, Lee C, Tiano H. Cyclooxygenase knockout mice. Models for elucidatingisoform-specific functions. Biochemical Pharmacology 1999;58:1237-46.

19. Kiefer JR, Pawlitz JL, Moreland KT, Stegeman RA, Hood WF, Gierse JK, Stevens AM, GoodwinDC, Rowlinson SW, Marnett LJ, Stallings WC, Kurumbail RG. Structural insights into the stereo-chemistry of the cyclooxygenase reaction. Nature 2000;405:97-101.

20. Bhattacharyya DK, Lecomte M, Rieke CJ, Garavito RM, Smith WL. Involvement of arginine 120,glutamate 524, and tyrosine 355 in the binding of arachidonate and 2-phenylpropionic acidinhibitors to the cyclooxygenase active site of ovine prostaglandin endoperoxide H synthase-1. JBiol Chem 1996;271:2179-84.

21. Kurumbail RG, Kiefer JR, Marnett LJ. Cyclooxygenase enzymes: catalysis and inhibition. CurrOpinion Struct Biol 2001;11:752-60.

22. Rainsford KD. Profile and mechanisms of gastrointestinal and other side effects of nonsteroidalanti-inflammatory drugs (NSAIDs). Am J Med 1999;107 (6A) 27S-36S.

23. Tilley SL, Coffman, Koller BH. Mixed messages: modulation of inflammation and immuneresponses by prostaglandins and thromboxanes. J Clin Invest 2001;108:15-23.

24. Narumiya S, Sugimoto Y, Ushikubi F. Prostanoid receptors: structures, properties, and functions.Physiol Rev 1999;79:1193-226.

25. Stock JL, Shinjo K, Burkhardt J, Roach M, Taniguchi K, Ishikawa T, Kim H-S, Flannery PJ,Coffman Tm, McNeish JD, Audoly LP. The prostaglandin E2 EP1 receptor mediates pain per-ception and regulates blood pressure. J Clin Invest 2000;107:325-31.

26. Patrignani P, Panara MR, Greco A, Fusco O, Natoli C, Iacobelli S, Cipollone F, Ganci A,Créminon C, Maclouf J, Patrono C. Biochemical and pharmacological characterization of thecyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J PharmacolExptl Ther 1994;271:1705-12.

27. Young JM, Panah S, Satchawatcharaphong C, Cheung PS. Human whole blood assays for inhi-bition of prostaglandin G/H synthases-1 and -2 using A23187 and lipopolysaccharide stimulationof thromboxane B2 production. Inflamm Res 1996;45:246-53.

28. Thomas MA, O'Grady GE, Swartz SL. Prostaglandin levels in human vitreous. Br J Ophthalmol1985;69:275-9.

29. Chang MS, Tsai JC, Yang R, DuBois RN, Breyer MD, O'Day DM. Induction of rabbit cyclooxyge-nase 2 in the anterior uvea following glaucoma filtration surgery. Curr Eye Res 1997;16:1147-51.

30. Kulkarni PS, Srinivasan BD. Cyclooxygenase and lipoxygenase pathways in anterior uvea andconjunctiva. In: Bito LZ, Stjernschantz J, editors. The ocular effects of prostaglandins and othereicosanoids. New York Alan R Liss 1989:39-52.

31. Graff G, Brady MT, Gamache DA, Spellman JM, Yanni JM. Transient loss of prostaglandin syn-thetic capacity in rabbit iris-ciliary body following anterior chamber paracentesis. Ocul ImmunolInflamm 1998;6:227-38.

32. Mizuno H, Sakamoto C, Matsuda K, Wada K, Uchida T, Noguchi H, Akamatsu T, Kasuga M.Induction of cyclooxygenase 2 in gastric mucosal lesions and its inhibition by the specific antag-onist delays healing in mice. Gastroenterolgy 1997;112:387-97.

33. Gilroy DW, Colville-Nash PR, Willis D, Chivers J, Paul-Clark MJ, Willoughby DA. Induciblecyclooxygenase may have anti-inflammatory properties. Nat Med 1999;5:698-701.

34. Morteau O, Morham SG, Sellon R, Dieleman LA, Langenbach R, Smithies O, Balfour Sartor R.Impaired mucosal defense to acute colonic injury in mice lacking cyclooxygenase-1 or cyclooxy-genase-2. J Clin Invest 2000;105:469-78.

41

General introduction

35. Gilroy DW, Tomlinson A, Willoughby DA. Differential effects of inhibitors of cyclooxygenase(cyclooxygenase 1 and cyclooxygenase 2) in acute inflammation. Eur J Pharmacol 1998;355:211-17.

36. Willoughby DA, Moore AR, Colville-Nash PR. COX-1, COX-2, and COX-3 and the future treat-ment of chronic inflammatory disease. Lancet 2000;355:646-8.

37. Masferrer JL, Kulkarni PS. Cyclooxygenase-2 inhibitors: a new approach to the therapy of ocu-lar inflammation. Surv Opthalomol 1997;41:S35-S40.

38. Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis B, Day R, Bosi Ferraz M,Hawkey CJ, Hochberg MC, Kvien TK, Schnitzer TJ, for the VIGOR study group. Comparison ofupper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. NEngl J Med 2000;343:1520-8.

39. Ray WA, Stein CM, Daugherty JR, Hall K, Arbogast PG, Griffin MR. COX-2 selective non-steroidal anti-inflammatory drugs and risk of serious coronary heart disease. Lancet2002;360:1071-3.

40. Jüni P, Rutjes AWS, Dieppe PA. Are selective COX 2 inhibitors superior to traditional nonsteroidal anti-inflammatory drugs? Adequate analysis of the CLASS trial indicates that this maynot be the case. BMJ 2002;324:1287-8.

41. Jones R. Efficacy and safety of COX 2 inhibitors. New data are encouraging but the risk:benefitratio remains unclear. BMJ 2002;325:607-8.

42. Cheng Y, Austin SC, Rocca B, Koller BH, Coffman TM, Grosser T, Lawson JA, FitzGerald GA. Roleof prostacyclin in the cardiovascular response to thromboxane A2. Science 2002;296:539-4.

43. Vane JR. Back to an aspirin a day? Science 2002;296:474-5.44. Nebert DW, Russell DW. Clinical importance of the cytochromes P450. Lancet 2002;360:1155-62.45. Skelly MM, Hawkey CJ. Potential alternatives to COX 2 inhibitors. New molecules may overtake

the COX 2 inhibitors debate. BMJ 2002;324:1289-90.

42

Chapter 1