JOURNAL OF OCULAR PHARMACOLOGYVolume 10, Number 4, 1994Mary Ann Liebert, Inc., Publishers

Intraocular Diclofenac andFlurbiprofen Concentrationsin Human Aqueous Humor

Following Topical ApplicationPHILIP P. ELLIS, DAVID S. PFOFF,

DUANE C. BLOEDOW, and MARIS RIEGEL

Department of Ophthalmology, University of Colorado School of Medicine, Denver, Colorado

ABSTRACT

Flurbiprofen Na and diclofenac Na, two ocular antiinflammatory agents, were investigated todetermine the aqueous humor concentrations in the human eye following topical application. Onehundred sixty-five patients undergoing cataract surgery received a single drop of either diclofenac Naor flurbiprofen Na at selected times prior to the surgical procedure. Aqueous humor samples were

aspirated at the beginning of surgery and a sensitive high-performance liquid Chromatographie assaywas used to determine the concentration of the antiinflammatory agent in the ocular fluid. Sampleswere obtained between 10 min and 24 hrs after a single instillation of the drug onto the cornea. Thehighest average concentration of diclofenac was 82 ng/ml at 2.4 hrs after instillation; concentrationsremained above 20 ng/ml for over 4 hrs. Thereafter, between 3 and 16 ng/ml diclofenac could beassayed through 24 hrs. The highest average concentration of flurbiprofen, 60 ng/ml, was found at2.0 hrs. The last detectable flurbiprofen concentration was measured at 7.25 hrs after instillation.

INTRODUCTION

In the past few years several topical ophthalmic preparations of non-steroidalantiinflammatory agents (NSAIDs) have been developed. They provide efficacy with little toxicity.These agents act primarily by blocking prostaglandin synthesis through inhibition of cyclo-oxygenase,the enzyme catalyzing the conversion of arachidonic acid to prostaglandins (1). Certain NSAIDsmay also limit the arachidonic acid available for leukotriene production (2). Flach has reviewed theclinical studies describing the therapeutic effectiveness of flurbiprofen, diclofenac and other NSAIDs;however, the clinical indications and usefulness of the various NSAIDs in ophthalmology are stillbeing defined (1).

This report describes the aqueous humor drug concentration-time profiles following a single,carefully administered dose. We have compared the concentrations of the two agents followingtopical administration and extended the sampling times in an attempt to follow the absorption andelimination of each of the drugs in the aqueous humor. Pharmacokinetic analyses of the averageaqueous humor concentrations for each drug were performed.

677

MATERIALS AND METHODS

Patients and Study DesignThe test group was drawn from patients scheduled for routine cataract surgery who were not

receiving other ophthalmic medications. Patients with preexisting ocular disease or complications atsurgery were excluded from the study. All patients consented to the procedure. Data from 165patients were included in the data analysis. There were 106 females and 59 males with an age rangeof 39-93 years. The mean age was 71.4 years. Patients were randomly assigned to either thediclofenac (n=98) or flurbiprofen (n=67) treatment group.

The commercial preparations of the drugs used in this study, diclofenac Na, 0.1%, (Volteren,Ciba Vision Ophthalmics, Atlanta, GA) and flurbiprofen Na, 0.03%, (Ocufen, AllerganPharmaceuticals, Irvine, CA) were diluted and assayed in triplicate to verify the concentrations of thelots used.

At selected times before the surgical procedure, a single drop (50 ul) of either topicalantiinflammatory product was instilled onto the cornea. A minimum of 5 min was allowed betweendiclofenac Na or flurbiprofen Na instillation and the administration of any other topical ocularmedication (e.g. antibiotic combination polymyxin B sulfate:neomycin sulfate: gramicidin B, or

cyclopentolate). Patients routinely received either the antibiotic combination or tobramycin drops theday before surgery; these were continued the day of surgery. To provide lid akinesia, a modified VanLint injection with 2% lidocaine with hyaluronidase was used. A superior and inferior peribulbarinjection of 0.75% bupivacaine with hyaluronidase completed the anesthesia. At the beginning ofcataract surgery, 0.1-0.2 ml of aqueous humor was aspirated through the limbus using a 1-mltuberculin syringe with a 26 gauge needle. Samples were frozen at -20°C until assayed.Drug Assay

Diclofenac and flurbiprofen were measured as the free acids using a high-performance liquidChromatographie technique developed in our laboratory for small volumes of aqueous humor (3).Briefly, the sample and internal standard were extracted and deproteinized using acetonitrile. Thesample was centrifuged and the supernatant dried under nitrogen. Following reconstitution in mobilephase the separation was performed on a C8 reverse-phase column using an acetic acid/acetonitrilemobile phase containing triethylamine. The wavelength of detection was 280 nm; the lower limit ofsensitivity was 0.3 ng on column for diclofenac and 0.4 ng flurbiprofen on column (3).Data Analysis

The data were plotted and means were calculated within grouped time intervals in order todetermine the overall pattern of the absorption and elimination of the drug from the aqueous humor.These mean curves were subjected to pharmacokinetic analysis using standard methods (4).

RESULTS

The commercial diclofenac Na preparation contained 112% of the labeled amount with a 50 uldose corresponding to 56.0 ug (52.0 ug free acid); commercial flurbiprofen Na contained 96% of thelabeled amount giving a dose of 14.4 ug (12.6 ug free acid).

Aqueous humor diclofenac and flurbiprofen concentrations versus time are depicted in thescattergrams (Figs. 1,2); each point is from a single patient. The line graphs showing averageconcentrations are superposed on the scattergrams to further characterize the range of values (Figs.1,2). The variability of the data is evident; however, between 10 and 25 min after topical dosing,consistent, measurable concentrations of diclofenac were found in the aqueous humor whileflurbiprofen was not detected until the 30 min sampling time (Fig. 1,2). It is apparent front the datadiclofenac also remains in the eye for a longer time than flurbiprofen. The pharmacokinetic analysisof diclofenac and flurbiprofen movement through the aqueous humor was based on the area under theaverage concentration vs time curve (AUC) for detectable concentrations (Figs. 1,2). Flurbiprofenconcentrations were not detectable after 7.25 hrs. For each group, the time course of drug movementthrough the aqueous humor is described by the mean residence time (4), that is, the average lifetimefor a drug molecule in the aqueous humor (Table 1). The scatter of the values did not allow moreextensive pharmacokinetic evaluation.

678

bfjtíucdtí<D

4hOO

3OStí tíw

oÜtícr

200

150

100 [

50

0 12 18 24

time, hr

Figure 1. Diclofenac concentration in human aqueous humor. The individual data points are includedwith the mean curve to characterize the range of values. The insert uses a log time scale to displaythe early values more clearly. Dose: 52.0 ug.

Mtítí"<U

4hO

.&lo1-1

t-loa.awtíootícrcd

300

250 h

200

150 h

100

0 12 18 24

time, hr

Figure 2. Flurbiprofen concentrations in human aqueous humor. The individual data points areincluded with the mean curve to characterize the range of values. The insert uses a log time scale todisplay the early values more clearly. Dose: 12.6 ug.

679

TABLE 1

Pharmacokinetic Data Derived from Aqueous Humor Diclofenac and Flubiprofen Concentrations.

Diclofenac FlubiprofenInstilled dose

Sodium salt, ug 56.0 14.4Free acid, ug 52.0 12.6

Aqueous humor concentration data (free acid)Average Measured Maxima

Average Cm«, ng/ml 82 60Tpe,k, hr 2.4 2.0AUC, (ng/ml) hr 384 186

Mean residence time, hr 7.4 3.7Relative bioavailability*

Flurbiprofen:Diclofenac 2.0

*(AUC nurb¡profen)/(AUCdiciofeiiac), corrected for dose difference

DISCUSSION

This study was designed to determine aqueous humor drug concentration-time profiles ofdiclofenac and flurbiprofen achieved after a single, carefully administered dose of each drug onto thecornea. The ocular bioavailability of antiinflammatory drugs depends primarily on their penetrationthrough the cornea. Diclofenac is slightly more lipophilic, (partition coefficient, 13.4 vs. 5.76),which is consistent with a faster rate of entry of the drug into the aqueous humor (5-7).

Ocular pharmacokinetic data on these antiinflammatory agents are sparse (8,9). Although thedata from the present study are variable, they do indicate that the average maximum aqueous humorconcentrations are similar (Cmax 82 ng/ml and 60 ng/ml for diclofenac and flurbiprofen respectively,Table 1) and occur near the same time (2.4-2.0 hr, Table 1). The data also indicate that diclofenacremains in the aqueous humor for a longer period of time than flurbiprofen (Figs. 1,2) as measured bythe mean residence time for the assayed drug within the aqueous humor (7.4 vs. 3.7 hr, Table 1).

Relative bioavilability was estimated from the area under the average aqueous humorconcentration curve, correcting for dose according to:

relative bioavailability =[dOSe diclofenac* AUCflurbiprofen]/[doSe flurbiprofen* AUCdiclofenac]-This calculation of relative bioavilability indicates that for equivalent doses, flurbiprofen provides for2.0 times greater drug exposure than does diclofenac. However, considering the doses actuallyadministered (12.6 ug flurbiprofen and 52 ug diclofenac), the aqueous humor drug exposure based onthe clinical doses is greater for diclofenac, (AUCfmrbiprofen/AUCdiciofenac = 0.48).

To our knowledge, previous studies in human aqueous humor have neither comparedconcentrations of diclofenac and flurbiprofen nor reported the time-concentration profiles following a

single dose of antiinflammatory agent (1,8,10-12, personal comm. J. Cheetham, Allergan Inc., IrvineCA). This may be due to the limited sensitivity of the assays employed.

Previous investigators have found a somewhat higher mean maximum concentration ofdiclofenac than we did, approximately 130 ng/ml vs. 82 ng/ml, (8,10,11); an average maximumdiclofenac concentration occurred at one hr vs 2.4 hrs observed in our study (8,11). Otherinvestigators found a flurbiprofen peak time of 90 min (personal comm. J. Cheetham, Allergan Inc.,Irvine, CA) vs the 2 hr peak time observed by us. These differences in the antiinflammatory profilesmay result from the instillation of multiple drops in the other studies. Ocular drug levels can beincreased with multiple instillations of topical drops when adequate time is allowed between dropadministrations. The peak time is not necessarily the same when determined after the final drop ofmedication in a multi-drop regimen rather than following a single instillation of the agent. With a

single instillation, we observed a biphasic rise in aqueous humor levels of flurbiprofen with the first

680

peak occurring at 85 min, and a fall in the mean prior to a second peak at 2 hr (Fig. 2). Thisphenomenon could be confounded by multiple instillations.

In this study we confirmed the entry of flurbiprofen into the aqueous humor at 30 min thatwas previously reported (personal comm. J. Cheetham, Allergan Inc., Irvine, CA). Otherinvestigators also described substantial variations in concentration values at specified sampling times(8,10-12). The wide range of values found at each sampling time for both antiinflammatory agents(Fig. 1,2), in other studies and in ours, may reflect inter-individual differences in penetration of thedrugs similar to that observed for flurbiprofen in rabbits (9). Using multiple sampling from a singlerabbit eye, Tang-Liu found a pronounced, but different, absorption pattern for each test animal, with aconcentration maximum in the aqueous humor at 2-4 hr (9).

Agata et al. measured the penetration of topically applied radioactive diclofenac in rabbit eyes;they found that corneal and aqueous humor diclofenac levels fell approximately in parallel (13). Thissuggests a diclofenac corneal depot similar to pilocarpine (14,15). Our data also support a cornealreservoir for diclofenac, with measurable concentrations observed throughout 24 hours.

The results of the present study are consistent with the observed clinical effects ofantiinflammatory agents and with their physicochemical properties (1,6-8). On the basis of the datafrom this study, one could speculate that diclofenac, which provides aqueous humor concentrationslonger lasting than flurbiprofen, would be a preferred agent for the treatment of post-operativeinflammation. Flurbiprofen, which provides greater short-term aqueous humor drug concentrationsthan diclofenac on an equivalent dose basis, is a well suited agent for the prevention of miosisoccurring during cataract surgery.

ACKNOWLEDGMENTS

This study was supported by an unrestricted grant from Research to Prevent Blindness. Theassistance of Theresa Heitung, R.N., in coordinating the collection of the samples is gratefullyacknowledged.

REFERENCES

1. Flach, A.J. Cyclo-oxygenase inhibitors in ophthalmology. Surv. Ophthalmol. 36:259-284,1992.

2. Cronin, ME. and Wortmann, R.L. Nonsteroidal antiinflammatory drugs (NSAIDs); commonchemical and clinical characteristics. Intl J. Dermatol. 23:411-413.

3. Riegel, M. and Ellis, P.P. High performance liquid Chromatographie assay forantiinflammatory drugs in the aqueous humor of the eye. J. Chromatog. 654:140-145, 1994.

4. Gibaldi, M. and Perrier, D. Pharmacokinetics., 2nd ed. New York, N.Y. Marcel Dekker, Inc,1982, p. 410.

5. van Haeringen, N.J., Oosterhuis, JA., van Delft. J.L., Glasius, E. and Noach, EL. Acomparison of the effects of non-steroidal compounds on the disruption of the blood-aqueousbarrier. Exp. Eve Res. 35:271-277. 1982.

6. Sallman, A. Chemical aspects of diclofenac. International Symposium for Polyarthritis,Torremolinos, 1975. In Chronic forms of Polyarthritis, Wagenhauser, F.J., ed., Hans Huber,Berne, 1976, p. 296-301

7. Risdall, P.C., Adams, S.S., Crampton, EL. and Marchant, B. The disposition and metabolismof flurbiprofen in several species including man. Xenobiotica 8:691-704, 1978.

8. Goa, K.L. and Chrisp, P. Ocular diclofenac. A review of its pharmacology and clinical use incataract surgery, and potential in other antiinflammatory conditions. Drugs Aging 2:473-486,1992.

9. Tang-Liu, D.D., Liu, S.S. and Weinkam, R.J. Ocular and systemic bioavailability ofophthalmic flurbiprofen. J. Pharmacokinet. Biopharm. 12:611-626. 1984.

681

10. Quentin, CD. Diclofenac-kammerwasserkonzentrationsbestimmung bei kataraktoperation.Deutsche Gesellschaft fur Intraokularlinsen Implantation, Aachen, G.D.R., March 8-9, 1991.In Kongress der Deutschen Gesellschaft fur Intraokularlinsen Implantation, Schott K, Jacobi,K.W., Freyler, H., eds., Springer-Verlag, Heidelburg, 1991.

11. Vickers, F.F, John, V.A., Powell, ML., Wysowskyj, H., Luders, R., Brunner, L. and Ahmed,M. Corneal penetration of 0.1% diclofenac sodium ophthalmic solutions in patientsundergoing cataract surgery. Annual Meeting of American College of Clinical Pharmacology,Las Vegas, NV, November 4-8, 1990. J. Clin. Pharm. 30:835, 1990 (Abstract).

12. Strobel, J. Kammerwasserkonszentrationen nichtsteroidaler antiphlogistika. Ophthalmologica,202:38-42, 1991.

13. Agata, M., Tanaka, M., Nakajima, A., Fujii, A., Kuboyama, N., Tamura, T., and Araie, M.Ocular penetration of topical diclofenac sodium, a non-steroidal anti-inflammatory drug, inrabbit eye. Nippon Ganka Gakkai Zasshi 88:61-66, 1984.

14. Makoid, M.C., and Robinson, J.R. Pharmacokinetics of topically applied pilocarpine in thealbino rabbit eye. J. Pharm. Sei. 68:435-443, 1979.

15. Grass, G.M, and Robinson, J.R. Relationship of chemical structure to corneal penetration andinfluence of low-viscosity solution on ocular bioavailability. J. Pharm. Sei. 73:1021-1027,1984.

Received: May 24, 1994Accepted for Publication: July 5, 1994

Reprint Requests: Philip P. Ellis, M.D.Department of OphthalmologyBox B-204University of Colorado Health Sciences Center4200 E. 9th AvenueDenver, Colorado 80262U.S.A.

682