Ophthalmic Drug Delivery Systems - part 5 doc

The main focus of this chapter will be on the use of penetration enhan-cers to improve ocular drug delivery.. 10 mM 6-Carboxyfluorescein Rabbit Enhanced penetrated amount 7.2 times2–10 mM

39 M Barza, A Kane, and J Baum (1980) Oxacillin for bacterial mitis: subconjunctival, intravenous, both or neither? Invest Ophthalmol Vis.Sci 19:1348–1354.

endophthal-40 W M Jay, R K Shockley, A M Aziz, M Z Aziz, and J P Rissing (1984).Ocular pharmacokinetics of ceftriaxone following subconjunctival injection inrabbits Arch Ophthalmol 102:430–432

41 F P Furgiuele, J P Smith, and J G Baron (1978) Tobramycin levels inhuman eyes Am J Ophthalmol 85:121–123

42 D M Maurice, and S Mishima (1984) Ocular pharmacokinetics In:Pharmacology of the Eye II M L Sears, ed New York: Springer-Verlag,

p 19

43 M Barza, A Kane, and J Baum (1983) Pharmacokinetics of intravitrealcarbenicillin, cefazolin, and gentamicin in rhesus monkeys Invest.Ophthalmol Vis Sci 24:1602–1606

44 V N Reddy, B Chakrapani, and C P Lim (1977) Blood-vitreous barrier toamino acids Exp Eye Res 25:543–545

45 A G Schenk and G A Peyman (1974) Lincomycin by direct intravitrealinjection in the treatment of experimental bacterial endophthalmitis AlbrechtVon Graefes Arch Klin Exp Ophthalmol 190: 281–291

46 S C Pflugfelder, E Hernandez, S J Fliesler, J Alvarez, M E Pflugfelder,and R K Forster (1987) Intravitreal vancomycin Retinal toxicity, clear-ance, and interaction with gentamicin Arch Ophthalmol 105:831–837

47 F M Ussery, 3rd, S R Gibson, R H Conklin, D F Piot, E W Stool, and

A J Conklin (1988) Intravitreal ganciclovir in the treatment of ciated cytomegalovirus retinitis Ophthalmology, 95:640–648

AIDS-asso-48 K Henry, H Cantrill, C Fletcher, B J Chinnock, and H H Balfour, Jr.(1987) Use of intravitreal ganciclovir (dihydroxy propoxymethyl guanine) forcytomegalovirus retinitis in a patient with AIDS Am J Ophthalmol 103:17–23

49 G A Peyman, D R May, E S Ericson, and D Apple (1974) Intraocularinjection of gentamicin Toxic effects of clearance Arch Ophthalmol 92:42–47

50 R K Shockley, W M Jay, T R Friberg, A M Aziz, J P Rissing, and M Z.Aziz (1984) Intravitreal ceftriaxone in a rabbit model Dose- and time-depen-dent toxic effects and pharmacokinetic analysis Arch Ophthalmol 102:1236–1238

51 A G Schenk, G A Peyman, and J T Paque (1974) The intravitreal use ofcarbenicillin (Geopen) for treatment of pseudomonas endophthalmitis ActaOphthalmol 52:707–717

52 S K Gardner (1987) Ocular drug penetration and pharmacokinetic ples In: Clinical Ophthalmic Pharmacology D W Lamberts and D E Potter,eds

princi-53 T S Lesar and R G Fiscella (1985) Antimicrobial drug delivery to the eye.Drug Intell Clin Pharm 19:642–654

54 J D Wright, F D Boudinot, and M R Ujhelyi (1996) Measurement andanalysis of unbound drug concentrations Clin Pharmacokinet 30:445–462

55 U Ungerstedt (1991) Microdialysis—principles and applications for studies

in animals and man J Intern Med 230:365–373

56 E C de Lange, M Danhof, A G de Boer, and D D Breimer (1997).Methodological considerations of intracerebral microdialysis in pharmacoki-netic studies on drug transport across the blood-brain barrier Brain Res.Brain Res Rev 25:27–49

57 R Tao and S Hjorth (1992) Differences in the in vitro and in vivo xytryptamine extraction performance among three common microdialysismembranes J Neurochem 59:1778–1785

5-hydro-58 J Landolt, H Langemann, T H Lutz, and O Gratzl (1991) Non-linearrecovery of cysteine and glutathione in microdialysis In: H Rollema, et al.,eds Meppel, Netherlands: Krips Repro, pp 63–65

59 J Waga, I Nilsson-Ehle, B Ljungberg, A Skarin, L Stahle, and B Ehinger.(1999) Microdialysis for pharmacokinetic studies of ceftazidime in rabbitvitreous J Ocul Pharmacol Ther 15:455–463

60 J Waga (2000) Ganciclovir delivery through an intravitreal microdialysisprobe in rabbit Acta Ophthalmol Scand 78:369–371

61 A Hamberger, C H Berthold, B Karlsson, A Lehmann, and B Nystrom.(1983) Extracellular GABA, glutamate and glutamine in vivo perfusion dia-lysis of the rabbit hyppocampus In: Glutamine, Glutamate and GABA in theCentral Nervous System New York: Alan R Liss Inc., pp 473–492

62 M Sandberg and S Lindstrom (1993) Amino acids in the dorsal lateralgeniculate nucleus of the cat—collection in vivo J Neurosci Methods 9:64–74

63 U Tossman and U Ungerstedt (1986) Microdialysis in the study of cellular levels of amino acids in the rat brain Acta Physiol Scand 128:9–14

extra-64 A Lehmann (1989) Effects of microdialysis-perfusion with anisoosmoticmedia on extracellular amino acids in the rat hippocampus and skeletal mus-cle J Neurochem 53:525–535

65 J M Solis, A S Herranz, O Herreras, J Lerma, and R Martin del Rio.(1988) Does taurine act as an osmoregulatory substance in the rat brain?Neurosci Lett 91:53–58

66 J V Wade, J P Olson, F E Samson, S R Nelson, T L Pazdernik (1988)

A possible role for taurine in osmoregulation within the brain J Neurochem.51:740–745

67 S A Wages, W H Church, and J B Justice, Jr (1986) Sampling tions for on-line microbore liquid chromatography of brain dialysate Anal.Chem 58:1649–1656

considera-68 J M Delgado, J Lerma, R Martin del Rio, and J M Solis (1984).Dialytrode technology and local profiles of amino acids in the awake catbrain J Neurochem 42:1218–1228

69 G Amberg, and N Lindefors (1989) Intracerebral microdialysis: II.Mathematical studies of diffusion kinetics J Pharmacol Methods 22:157–183

70 T Zetterstrom, L Vernet, U Ungerstedt, U Tossman, B Jonzon, and B B.Fredholm (1982) Purine levels in the intact rat brain Studies with animplanted perfused hollow fibre Neurosci Lett 29:111–115

71 I Jacobson, M Sandberg, and A Hamberger (1985) Mass transfer in braindialysis devices—a new method for the estimation of extracellular amino acidsconcentration J Neurosci Methods 15:263–268.

72 P Lonnroth, P A Jansson, and U Smith (1987) A microdialysis methodallowing characterization of intercellular water space in humans Am J.Physiol 253:E228–231

73 D Scheller and J Kolb (1991) The internal reference technique in dialysis: a practical approach to monitoring dialysis efficiency and to calculat-ing tissue concentration from dialysate samples J Neurosci Methods 40:31–38

micro-74 H Benveniste, A J Hansen, and N S Ottosen (1989) Determination ofbrain interstitial concentrations by microdialysis J Neurochem 52:1741–1750

75 G Raviola (1974) Effects of paracentesis on the blood-aqueous barrier: anelectron microscope study on Macaca mulatta using horseradish peroxidase as

a tracer Invest Ophthalmol 13:828–858

76 F P Killey, H F Edelhauser, and T A Aaberg (1980) Intraocular fluiddynamics Measurements following vitrectomy and intraocular sulfur hexa-fluoride administration Arch Ophthalmol 98:1448–1452

77 L L Knudsen, T Olsen, and F Nielsen-Kudsk (1988) Anterior chamberfluorescein kinetics compared with vitreous kinetics in normal subjects ActaOphthalmol Scand 76:561–567

78 J Wagar and B Ehinger (2000) NGF administered by microdialysis intorabbit vitreous Acta Ophthalmol Scand 78:154–155

Ocular Penetration Enhancers

Thomas Wai-Yip Lee and Joseph R Robinson

School of Pharmacy, University of Wisconsin-Madison, Madison,

a challenge for pharmaceutical scientists to design drug delivery systems thatcan deliver therapeutic agents in sufficient concentrations to target sites.After topical instillation of an eye drop, the drug is subject to a num-ber of very efficient elimination mechanisms such as drainage, binding toproteins, normal tear turnover, induced tear production, and nonproductiveabsorption via the conjunctiva Typically, drug absorption is virtually com-plete in 90 seconds due to the rapid removal of drug from the precornealarea To make matters worse, the cornea is poorly permeable to both hydro-philic and hydrophobic compounds As a result, only approximately 10% orless of the topically applied dose can be absorbed into the anterior segment

drugs, penetration enhancers, etc Prodrugs will be covered elsewhere in thisbook The main focus of this chapter will be on the use of penetration enhan-cers to improve ocular drug delivery Fundamental aspects of ocular penetra-tion enhancers will be covered, and recent advances will be presented as well.

ENHANCEMENT

The simplest model for ocular pharmacokinetics is shown in Figure 1 (5) It

is well known that for most drugs the true absorption rate constant is muchsmaller than the elimination rate constant This will normally give rise to aflip-flop model However, when the parallel elimination pathway is intro-duced (Fig 2) (5), the apparent absorption rate constant is defined as:Apparent kabs¼ kabsþ kloss;pp

Thus, the model is not a flip-flop model and drug concentration can bedescribed as

where F is the fraction of dose absorbed, D is the dose, k and K areabsorption and elimination rate constants, respectively, and Vd is the appar-ent volume of distribution Obviously, K ¼ kelim, k ¼ kabsþ kloss;pp:For many drugs, kloss; ppis of the order of 0.5–0.7 min1, being severalorders of magnitude larger than kabs, which is typically of the order of 0.001min1 As a result, the peak time, which is controlled by kloss;pp and kabs, issimilar (20–40 min) for a wide range of compounds since kloss ;pp, which ismainly due to drainage, induced lacrimination, etc., predominates over kabs

in controlling the peak time

In order to improve the bioavailabilityðF ¼ kabs=½kabsþ kloss ;ppÞ nificantly, it is essential to increase kabsby one or two order of magnitudes

sig-or reduce kloss ;ppto a similar extent.

Several approaches have attempted to reduce the magnitude of kloss ;pp.However, it has its limit Keister et al (6) showed that reducing the dosevolume from 25 mL to zero brings only a fourfold improvement in bioavail-ability for a poorly permeable compound However, it is practically impos-

Figure 1 A one-compartment model for ocular absorption

diffusion, and carrier-mediated transport In contrast, the latter representsdiffusive and convective transport occurring through intercellular spacesand tight junctions Due to its aqueous nature, hydrophilic solutes wouldpreferably adopt the paracellular pathway However, there are three forms

of junctional complexes that form between cells which hinder transport ofhydrophilic molecules, namely, tight junctions (zonula occludens), inter-mediate junctions (belt desmosome or zonula adherens), and spot desmo-somes (macula adherens) (Fig 3) (9) Among them, the tight junction is theuppermost and tightest, and it gives the greatest resistance for hydrophilicmolecules to go between cells The barrier property of the tight junction can

be reflected by the transepithelial electrical resistance (TEER) The higherthe TEER, the tighter the junctions that give a higher resistance for trans-port of molecules Generally, epithelia with resistances in the range of 10–

compar-The cornea also shows permselectivity (11) It has an isoelectric point(pI) of 3.2 At pHs above the pI, it carries a negative charge and is selective

to positively charged molecules On the other hand, at pHs below the pI, it

Table 1 Expected Mechanisms of Corneal Penetration

Drug type

Apparent rate-limiting

Water soluble Epithelium Low o/w partition into epithelium

Slow diffusion through epitheliumHigh partition rate + rapiddiffusion through stroma/endothelium

Via leaky channelsSolute movement may beintercellular and/or transcellularWater and oil

Ocular Penetration Enhancers 287

Figure 4 A simplified diagram of histology of the cornea (Modified from Ref 8.)

The innermost layer is the endothelium Although the endothelium is philic, it is leaky and does not give any significant resistance to the transport

lipo-of molecules It is believed that the epithelium provides the major resistancefor hydrophilic/charged molecules and gives minimal resistance to smalllipophilic molecules However, after passing across the epithelium, furthermovement of these lipophilic molecules is limited by the matrix, which ishydrophilic in nature As a result, in order to pass across the whole cornea,the molecule has to have a balance between its lipophilic and hydrophiliccharacter

Other transport mechanisms such as carrier-mediated transport, cytosis, etc may also be involved in transcellular transport but they arepoorly understood

4 The enhancers should be physically and chemically compatiblewith a wide range of drugs and excipients.

However, currently available penetration enhancers are far from satisfyingthe above requirements None have yet been approved by the FDA presum-ably because of safety concerns In order to design an efficient and safepenetration enhancer, it is necessary to have a thorough understanding ofthe mechanisms of penetration enhancement Basically, penetration enhan-cers work by one or more of the following mechanisms (13):

1 Altering membrane structure and enhancing transcellular port by extracting membrane components and/or increasingfluidity

trans-2 Enhancing paracellular transport:

Chelating calcium ions leads to opening of tight junctions;Inducing high osmotic pressure that transiently opens tightjunctions;

Introducing agents to disrupt the structure of tight junctions

3 Altering mucus structure and rheology so that this diffusion rier is weakened

bar-4 Modifying the physical properties of the drug-enhancer entity

5 Inhibiting enzyme activity

A summary of ocular penetration enhancers is shown in Table 4 (14).Typically, ocular penetration enhancement falls into two categories: para-cellular and transcellular

A Enhanced Paracellular Transport

As mentioned earlier, tight junctions are the major determinant of lular transport In other words, tight junctions are the primary targets for apenetration enhancer to act on in order to improve paracellular transport.The most well-known penetration enhancer to improve paracellular trans-port is EDTA, which is a calcium chelator commonly used as a preservative

paracel-It is well known that proper functioning of tight junctions depends oncalcium ions In the absence of calcium ions, there is a widening of tightjunctions, resulting in an increase in paracellular permeability (8) EDTAcan remove divalent ions by its chelating action Therefore, there is nosurprise that it has a permeabilizing effect on biological membranes (16).However, its action on the cornea is believed to be much more complicated.Rojanasakul et al (17) showed that severe membrane damage is evident incorneas treated with EDTA, bile salts, and surfactants This disruption ofplasma membrane structures by EDTA is somewhat unexpected since it isbelieved that EDTA only interferes with the ability of calcium to maintain

10 mM 6-Carboxyfluorescein Rabbit Enhanced penetrated amount 7.2 times2–10 mM FD-4 Rabbit Enhanced penetrated amount slightlyTaurodeoxycholic acid 0.05% Atenolol, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 5.8 times for atenolol and

1.6 times for timolol0.075–0.1% Timolol Rabbit Enhanced Papp 5.2–5.5 times

10 mM 6-Carboxyfluorescein Rabbit Enhanced penetrated amount 593 times2–10 mM FD-4 Rabbit Enhanced penetrated amount 30.9–61.5

timesUrodeoxycholic acid 0.05% Atenolol, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 2.1 times for timolol and

1.6 times for betaxolol0.075–0.1% Timolol Rabbit Enhanced Papp 8.3–11.0 timesTauroursodeoxycholic

acid

0.05% Atenolol, Timolol,

Levobunol,Betaxolol

Rabbit Enhanced Papp 3.0 times for atenolol and

1.5 times for betaxolol0.075–0.1% Timolol Rabbit Enhanced Papp 3.3 times at 0.1%

Fatty acidsCapric acid 0.5% Atenolol, Carteolol,

Tilisolol, Timolol,Befunolol

Rabbit Enhanced Papp 20.3 times for atenolol,

8.9 times for carteolol, 5.1 times fortilisolol, and 3.0 times for timololTable 4 Continued

PreservativesBenzalkoniumchloride

0.01% Prostaglandin F2a,

Pilocarpine,Dexamethasone

Pig Enhanced Papp 7.2 times for

prostaglandin F2a, 1.7 times forpilocarpine, and 3.3 times fordexamethasone

0.01% Tilisolol, 4,

FD-10

Rabbit Enhanced Papp 3.5 times for tilisolol, 28.8

times for 4, and 37.1 times for 10

FD-0.02% Atenolol, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 5.2 times for atenolol, 2.7

times for timolol, and 1.3 times forbetaxolol

0.05% FD-4, FD-10 Rabbit Enhanced Papp 43.6 times for FDA and

60.6 times for FD-100.005-0.02% Fluorescein Rabbit Increased permeability 4–2.5 times 0.02%0.01–0.03% Carbachol Rabbit Enhanced miotic response about 20 times0.025% Titmolol Rabbit Enhanced the ocular absorption about

80% and the systemic absorption about40%

Chlorhexidinedigluconate

0.01% Pilocarpine,

Dexamethasone

Pig Enhanced Papp 1.5 times for

dexamethasone0.0025–0.05% Fluorescein Rabbit,

Dexamethasone

Pig Enhanced Papp 1.8 times for pilocarpine

and 4.7 times for dexamethasone

Rabbit Enhanced Papp 2.7 times for tilisolol, 5.6

times for FD-4, and 4.8 times for FD-10Paraben 0.04% Tilisolol, FD-4, FD-

10

Rabbit Enhanced Papp 1.9 times for FD-10

Propyl paraben 0.02% Dexamethasone Pig Enhanced Papp 1.5 timesChelating Agents

EDTA 0.5% Atenolol, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 1.4 times for atenolol

0.5% Atenolol, Carteolol,

Tilisolol, Timolol,Befunolol

Rabbit Enhanced Papp 1.7 times for atenolol, 2.9

times for carteolol, 2.3 times for timolol,and 1.6 times for befunolol

0.5% FD-4, FD-10 Rabbit Enhanced Papp 15.5 times for FDA and

39.0 times for FD-100.1–0.5% Atenlool, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 31 times for atenolol and

1.9 times for timolol at 0.5%

0.05% Timolol Rabbit Enhanced ocular and systemic absorption

significantlyOthers

Azone 0.025–1.0% Cimetidine Rabbit Enhanced Papp 14.1–87.0 timesTable 4 Continued

Rabit Enhanced Papp 29.1 times for

acetazolamide, 16.3 times forguanethidine, >87.3 times forguanethidine, 31.3 times for cimetidine,2.2 times for bunolol, and 2.2 times forprednisolone

0.025–0.1% Cimetidine Rabbit Enhanced ocular bioavailability 3.9–22.0

times after instillation in rabbits5% Cyclosporine Rabbit Enhanced penetration into the cornea and

rapidly achieved state-steady state druglevel

HexamethyleneLauramide

0.025–1.0% Cimetidine Rabbit Enhanced Papp 17.4-64.3 timesHexamethylene

Octanamide

0.025–1.0% Cimetidine Rabbit Enhanced Papp 5.7–100.3 timesDecylmethylsulfoxide 0.025–1.0% Cimetidine Rabbit Enhanced Papp 25–77 timesSaponin 0.05% Atenolol, Timolol,

Levobunolol,Betaxolol

Rabbit Enhanced Papp 16.5 times for atenolol,

11.0 times for timolol, 1.3 times forlevobunolol, 2.0 times for betaxolol0.01–0.025% Timolol Rabbit Enhanced Papp 2.1 times at 0.01%, 3.3

times at 0.015%, and 8.3 times at0.025%

Rabbit Enhanced Papp 31.9 times for atenolol,

13.2 times for carteolol, 7.6 times fortilisolol, 3.3 times for timolol, 2.7 timesfor befunolol

0.5% FD-4, FD-10 Rabbit Enhanced Papp 100 times for FD-4 and

intercellular integrity, but the effect may be due to concentration and tact time Nishihata et al (18) showed that EDTA caused leakage of cellproteins from rectal epithelia Therefore, there is a possibility that EDTAcan exert multiple effects on biological membranes However, it is believedthat its primary action is still on the integrity of tight junctions since it fails

con-to improve delivery of progesterone, which appears con-to penetrate the corneaprimarily by the transcellular route (16)

Cytochalasins are a group of small molecules that bind specifically toactin microfilaments, the major component of the cytoskeleton It has beenshown that the cytoskeleton participates in regulation of epithelial perme-ability in a variety of conditions (19) Therefore, it is a reasonable strategy todesign a penetration enhancer to act specifically on the cytoskeleton in order

to improve paracellular transport Rojanaskul et al (17) showed that chalasin B decreases TEER of the cornea in a dose-dependent manner Theyalso studied the safety profile of cytochalasin B in vitro Confocal micro-scopy showed that cytochalasin B produced negligible damage effect on thecell membrane Moreover, replacement of cytochalasin B after 30-minutetreatment with drug-free GBR results in a complete restoration of TEER, aprocess that is completed within 30 minutes after solution replacement.However, prolonged exposure time (e.g., > 1 hour) results in permanentdamage with incomplete recovery within the time frame of the experiment

cyto-It is obvious that cytochalasin B is a relatively specific and safe ocularpenetration enhancer compared with other classical penetration enhancerssuch as bile salts, surfactants, etc

Another strategy to improve paracellular transport is to make use ofactive transport systems Active transport of glucose or amino acids, which

is coupled to sodium transport, across the intestinal mucosa into the cellular lateral spaces creates an osmotic force for fluid flow, and this inturns triggers contraction of the perijunctional actomyosin, resulting inincreased paracellular permeability (decreased TEER) (20) Martinez-Palomo (21) showed that a hypertonic lysine solution induced a reversibleopening of the tight junction of the toad urinary bladder without grossdeformation of tight junctions The decreased TEER was reversed when

inter-an isotonic solution was replaced on the apical side of the epithelium.Due to complete reversibility, it appears that increased paracellular trans-port by applying a hypertonic solution works in isolated tissues However, ahypertonic solution may irritate the eye and induce tear production, whichflushes away the applied drug Therefore, the practicality of this approach isquestionable and has yet to be confirmed

B Enhanced Transcellular Transport

Enhancers that increase transcellular permeability to drugs probably do so

by affecting membrane lipids and protein components It is shown that fattyacids and their derivatives have been found to act primarily on the phos-pholipid component of membranes thereby creating disorder, resulting inincreased permeability (22)

Membrane cholesterol is another target for enhanced transcellulardelivery It was postulated that extraction of cholesterol out of the epithelialmembrane by medium chain monoglycerides, glyceryl-monooctanoate, gly-ceryl-1-monodecanoate, and glyceryl-monododecanoate promoted rectalcefoxitin absorption However, some fatty acids act on the protein compo-nent in membranes This is certainly the case for caprylate (22)

Nonprotein thios are another membrane component where certainenhancers can act The good correlation between reduced nonprotein thiolsand enhanced transport of hydrophilic compounds suggests an importantrole for nonprotein thiols in preventing the transport of hydrophilic com-pounds (22) Murakami et al (23) showed that depletion of these nonpro-tein thiols by treating with SH-modifying agents like diethyl maleate, diethylethoxymethylenemalonate, ethanol, or alicylates enhanced the mucosal toserosal transport of many hydrophilic compounds including cefoxitin andphenol red in rat intestinal tissue

Typically, classical penetration enhancers have a nonspecific action on logical membranes They work by reversibly or permanently damaging themembranes so that their safety is questionable Newer penetration enhan-cers have been introduced in ocular drug delivery recently with an aim tosolving this problem

bio-A Cyclodextrin

Cyclodextrins are a group of homologous cyclic oligosaccharides consisting

of six, seven, and eight glucose units, namely, a-, b-, and g-cyclodextrin (Fig.5) (24), respectively Typically, cyclodextrins act as true carriers by keepinghydrophobic molecules in solution by their hydrophobic cores, i.e., com-plexation between hydrophobic molecules and the inner hydrophobic core

of cyclodextrins In other words, they are not capable of modifying thepermeability of a biological barrier On the other hand, drug absorptionmay be limited by the release of the drug from the drug-cyclodextrin com-

Loosening of tight junctions and subsequent water (corneal swelling wasobserved in the study) and drug influx might also be one of the possiblemechanisms It was believed that retardation of absorption of lipophilicmolecules was due to creation of a more hydrated barrier, which retardsentry of lipophilic molecules However, this is a pure speculation, andfurther experiments need to be done to clarify the exact mechanism.Safety is still the major concern in using Azone1 as an ocular pene-tration enhancer It is necessary to keep the Azone1 concentration to aminimum (<0.1%) since higher concentrations might cause ocular discom-fort, conjunctival hyperemia, and epithelial thinning as a result of erosionand/or atrophy (33) Fortunately, in vitro experiments showed that 0.1%was enough to enhance penetration for most of the compounds under inves-tigation to a significant extent (33).

Saponin is a type of polysaccharide isolated from the bark of the Quillajasaponaria tree Saponin is an amphiphilic compound that has surface activ-ity As a result, although the exact mechanism of the penetration enhancingeffect is not well understood, it is believed that the penetration enhancingeffect relies solely on its detergent action 0.5% Saponin increased cornealpermeability of atenolol by two- to threefold but only slightly improved thedelivery of relatively lipophilic timolol and befunolol Saponin was alsoextensively studied as a penetration promoter for systemic delivery ofmacromolecules via the eye Saponin demonstrated improvement in thesystemic delivery of insulin and glucagon via the ocular route in variousspecies (34–39) It was shown that the penetration-enhancing effect of sapo-nin was simply a direct detergent action since there did not exist a linearcorrelation between efficacy and surfactant strength Therefore, the exactmechanism is still a mystery The major concern in using saponin in anophthalmic product is still safety Concentrations higher than 0.5% areirritating to the eye (35)

A a-Amino Acid

Most of classical penetration enhancers improve the paracellular transportacross a biological membrane by damaging the tight junctions to varyingdegrees by their nonspecific actions As a result, few of them are approved

by the FDA because of safety concern Emisphere Technologies synthesized

a series of small molecular weight a-amino acids, which are used to promote

oral delivery (Fig 6) (40,41) These delivery agents successfully increasedabsorption of several macromolecules in vivo in rats and primates, includinghumans, such as salmon calcitonin (42), interferon-a (43), heparin (44), andhuman growth hormone (hgH) (45) Wu (41) showed that these carriers canincrease the permeability coefficient of human growth hormone acrossCaco-2 monolayers by 10-fold Although it did have some effects on para-cellular transport, the major pathway was observed to be transcellular.Failure of these carriers to improve the transport of hydrocortisone, atranscellular marker, in Caco-2 monolayers showed that these carriershave a specific interaction with hGH, which makes the hGH more trans-portable, and such an interaction does not exist in the case of hydrocorti-sone It was clearly established that these carriers do not damage cellmembranes and thus are not classical penetration enhancers Moreover,the carrier-drug complex is not absorbed by an active transport process.

Figure 6 Chemical structures of various amino acid derivative carriers (From Ref.41.)

The working mechanistic assumption is that the carrier shields those philic groups on the molecule that restrain absorption.

hydro-Previous work in our laboratory showed that these carriers increasethe permeability coefficient of hGH across the cornea of a rabbit by 10-fold(46) Further study is ongoing in our laboratory to confirm the efficacy andtoxicity of these carriers as delivery agents/carriers for ocular drug delivery

The enhancement effect of Pz-peptide on permeability of drug acrossthe cornea and conjunctiva was studied by Chung et al (49) Pz-peptideincreases penetration across the cornea and conjunctiva for a wide range ofcompounds such as atenolol, propranolol, mannitol, fluorescein, FITC-Dextran 4000, etc Compared with other traditional penetration enhancerssuch as a cytochalasin B and EDTA, Pz-peptide is less potent in facilitatingparacellular transport since it fails to improve the penetration of FITC-Dextran 10000 across the cornea

The mechanism of enhancement is believed to involve stimulation oftransepithelial Na+flux at the level of the amiloride-sensitive Na+channeland then triggering biochemical changes, which result in opening of tightjunctions This was demonstrated in colonic segments of rabbits and Caco-2cell monolayers (50) However, this may not be the case in ocular tissues.Amiloride (a Na+ channel blocker), hexamethylene amiloride (Na+/H+exchange blocker), ouabain (a Na+/K+ ATPase inhibitor), and replace-ment of Na+ with choline chloride fails to inhibit Pz-peptide penetration(49) In addition, Pz-peptide can unexpectedly enhance the penetration ofpropranolol, which transports across a biological membrane solely via atranscellular pathway This may be due to inhibition of Pgp170 drug effluxpump, which was found in the conjunctiva (51), since propranolol is asubstrate for this efflux system Further investigation has to be carried out

to clarify its exact mechanism of enhancement

Although the enhancement effect of Pz-peptide is promising in vitro(2.4–5.1 for cornea, 1.8–2.0 for the conjunctiva in the case of atenololand propranolol), its effect is much less pronounced in vivo Pz-peptide fails

to enhance ocular absorption of propranolol and only improves the tion of atenolol by 1.4–2.0 times This may be due to the dilution effect of

absorp-resident tears on Pz-peptide and the applied dose or binding of Pz-peptide tomucin in other tears proteins.

C Multifunctional Approach (Polymeric Penetration

Enhancers)

1 Colloidal Systems

Colloidal systems have been extensively studied as carriers for ocular drugdelivery (52) The mechanism of enhancement is generally believed to berelated to prolonged residence time in the cul-de-sac However, enhancedpenetration may also be one of the explanations for improved ocular deliv-ery Poly-e-caprolactone nanoparticles, nanocapsules, and submicron emul-sions improved ocular bioavailability of indomethacin when compared withaqueous solutions and with a suspension of microparticles (53) It is believedthat the colloidal nature, rather than the inner structure or the specificcomposition of the colloidal carriers, plays a key role in the enhancementsince all three colloidal carriers improve the ocular bioavailability of indo-methacin to a similar extent Confocal microscopy showed that the colloidalcarriers penetrate into the epithelial cells of the cornea without causingdamage to the cell membrane This suggests that these carriers enter theepithelium via endocytosis Therefore, these carriers act as a penetrationenhancer or an endocytotic stimulator

2 Bioadhesives

A mentioned earlier, in order to improve ocular bioavailability, either kelior

kabs have to be increased two- to threefold Various means have beenattempted to prolong residence time (1–4) Obviously, it is desirable tohave a delivery system that can stay in the precorneal area for an extendedperiod of time but at the same time enhance corneal penetration A number

of bioadhesive polymers have such properties Typically, these are molecules that have already been approved by FDA for other purposes.Therefore, safety should not be a big problem

macro-3 Polyacrylates

Poly(acrylic acid) derivatives such as Carbomer and Polycarbol are usedextensively as bioadhesives (54) They do have a membrane-penetratingenhancing effect, although the exact mechanism is not well understood Itwas demonstrated that polyacrylic acid gel significantly increased the influx

of water in the rat rectum (55) It was speculated that this solvent drag wasresponsible for the enhanced absorption of low molecular weight com-

pounds Reduction of mucus on the microvillus and dilatation of the cellular space was also observed 5–10 minutes after administration of poly-acrylic acid gel into the rat rectum However, these changes were reversibleand returned to normal relatively soon It was not likely that polyacrylicacid gel enhanced penetration solely by its detergent action since it inhibitedrather than induced hemolysis, which is commonly observed with surfac-tants Another possible mechanism is related to its chelating activity.Polycarbopol and other polyacrylic acid–based polymers are able to chelatecalcium (56), which is an essential component for proper functioning of tightjunctions In addition, chelation of cations that are essential for normalactivity of enzymes further improves bioavailability However, this inhibi-tory effect may be too weak to account for the improved bioavailability (57).

inter-In the case of ocular drug delivery, there is reported to be only a minimumamount of metabolizing enzymes in the precorneal area As a result, there isnot likely to be a drastic improvement in ocular bioavailability because ofthis enzyme inhibitory effect

4 Chitosan and Derivatives

Chitosan (poly[b-(1-4)-2-amino-2-deoxy-D-glucopyranose]) (58) is a philic, biocompatible, biodegradable polymer of low toxicity It is widelyused as a pharmaceutical excipient for direct compression of tablets, con-trolled release rate of drugs from a dosage form, enhanced dissolution, etc.(58) It also shows strong mucoadhesive properties (59) Chitosan was eval-uated as a delivery system to increase precorneal drug residence times (60).The positively charged chitosan can reduce the elimination rate from theprecorneal area by increasing viscosity and by its interaction with negativelycharged mucus (mucoadhesive) The presence of chitosan with tobramycincan bring an improvement in AUC and t1 =2 in the precorneal area.Moreover, this preparation is well tolerated with minimal toxicity

hydro-Besides increasing residence time of a drug in the precorneal area,chitosan can be used as a potential penetration enhancer to improve deliveryacross the cornea Dodane et al (61) showed that chitosan caused a rever-sible, time and dose-dependent decrease in TEER in Caco-2 cell monolayers.The increase in permeability was further confirmed by increased mannitolpermeability They suggested that the above effects might be due to partialalteration of the cytoskeleton, but the exact mechanism is not known Theslight perturbation of the plasma membrane was evident by the rise inextracellular LDH release However, complete recovery was observed 24hours after exposure to a low concentration of chitosan (0.005%) for ashort period of time (<60 min) Moreover, the chitosan did not affect cellviability as shown by the trypan blue exclusion test

Conjugation of an enzyme inhibitor to chitosan is another strategy toimprove drug delivery (62,63) However, due to the limited amounts ofenzymes in the precorneal area, this strategy may not be beneficial in oculardrug delivery.

It was shown that mucus may inhibit the binding of chitosan to anepithelia surface and hence decreases its absorption-enhancing effect inintestinal epithelia The absorption-enhancing effect of chitosan has notbeen evaluated in the eye The extent of inhibition of chitosan binding tomucus in the precorneal area has yet to be determined

The use of penetration enhancers in ocular drug delivery has been studiedfor more than a decade, but none has been approved by the FDA mainlydue to their nonspecific actions, which often give an unfavorable safetyprofile In order to design a more specific penetration enhancer, it is neces-sary to have a better understanding of membrane transport, physiology oftight junctions, etc Another alternative is to reversibly modify physico-chemical properties of a drug so that it becomes more transportable (e.g.,prodrugs or carriers)

Obviously, penetration enhancement has its limit It is not possible toincrease drug bioavailability indefinitely by use of penetration enhancementalone Other approaches such as increased residence time and inhibition ofmetabolizing enzymes should be used in conjunction with penetrationenhancement We hope that the coming biomaterial era will bring us such

a drug delivery system

REFERENCES

1 Sieg, J W., and Robinson, J R Vehicle effects on ocular drug bioavailability

1 Evaluation of fluorometholone J Pharm Sci 64:931–936, 1975

2 Patton, T F., and Robinson, J R., Ocular evaluation of polyvinyl alcoholvehicle in rabbits J Pharm Sci 64:1312–1316, 1975

3 Robinson, J R., and Mlynek, G M Bioadhesive and phase-change polymersfor ocular drug delivery Adv Drug Delivery Rev 16:45–50, 1995

4 Salminen, L., and Urtti, H Prolonged pulse-entry of pilocarpine with a ble drug insert Graefes Arch Clin Exp Ophthalmol 221:96, 1983

solu-5 Worakul, N., and Robinson, J R Ocular namics Eur J Pharm Biopharm 44:71–83, 1997

6 Keister, J C., Cooper, E R., Missel, P J., Lang, J C., and Hager, D F.Limits on optimizing ocular drug delivery J Pharm Sci 80:50–53, 1990.

7 Harris, D., and Robinson, J R Bioadhesive polymers in peptide drug ery Biomaterials 11:652–658, 1990

deliv-8 Grass, G M Mechanisms of Corneal Drug Penetration Ph.D thesis, School

of Pharmacy, University of Wisconsin–Madison, 1985

9 Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., and Watson, J D.Molecular Biology of the Cell, 3rd ed., Garland Publishing, Inc., New York,1994

10 Liaw, J., and Robinson, J R Ocular penetration enhancers In: OphthalmicDrug Delivery Systems (Mitra A., ed.), Marcel Dekker, New York, 1993, pp.369–381

11 Rojanasakul, Y., and Robinson, J R Transport mechanisms of the cornea:Characterization of barrier permselectivity Int J Pharm 55:237–246, 1989

12 Lee, V H L., and Robinson, J R Review: Topical ocular drug delivery:Recent developments and future challenges J Ocular Pharmacol 2:67–108,1986

13 Robinson, J R., and Yang, X Absorption enhancers In: Encyclopedia ofPharmaceutical Technology, Vol 18 (Swarbrick J and Boylan J C (eds.),Marcel Dekker, New York, 1999, pp 1–27

14 Sasaki, H., Yamamura, K., Mukai, Nishida, K., Nakamura, J., Nakashima,M., and Ichikawa, M Enhancement of ocular drug penetration, Crit Rev.Drug Carrier Syst 16:85–146, 1999

15 Junginger, H E., and Verhoef, J C Macromolecules as safe penetrationenhancers for hydrophilic drugs—a fiction? Pharmaceut Sci Technol.Today 1:370–376, 1998

16 Grass, G M., Wood, R W., and Robinson, J R Effects of calcium chelatingagents on corneal permeability Invest Ophthalmol Vis Sci 26:110–113,1985

17 Rojanasakul, Y., Liaw, J and Robinson, J R Mechanisms of action of somepenetration enhancers in the cornea: laser scanning confocal microscopic andelectrophysiological studies Int J Pharm 66:131–142, 1990

18 Nishihata, T., Tomida, H., Frederick, G., Rytting, J H., and Higuchi, T.Comparison of the effects of sodium salicylate, disodium ethylenediaminetetraacetic acid and polyoxylene-23-lauryl ether as adjuvants for the rectalabsorption of sodium cefoxitin J Pharm Sci 37:159, 1985

19 Frederiksen, O., and Leyssac, P P Effects of cytochalasin B and sulsulfoxide on isoosmotic transport by rabbit gall-bladder in vitro J Facial(Lund.) 265:103–118, 1977

dimethyl-20 Madara, J L., et al Effects of cytochalasin D on occluding junctions ofintestinal absorptive cells: Further evidence that the cytoskeleton may influ-ence paracellular permeability and junctional charge selectivity J Cell Biol.102:2125, 1986

21 Martinez-Palomo, A Structure of tight junctions in epithelia with differentpermeability Proc Natl Acad Sci USA 72:4487–4491, 1975

22 Lee, V H L., ed Peptides and Proteins Drug Delivery Marcel Dekker, NewYork, 1990.

23 Murakami, M., et al Intestinal absorption enhanced by unsaturated fattyacids: Inhibitory effect of sulfhydryl modifiers Biochim Biophys Acta293:238, 1988

24 Cyclodextrins in Pharmacy (Fromming, K.-H., and Szejtli, J., eds.), KluwerAcademic Publishers, Boston, 1994

25 Davies, N M., Wang, G., and Tucker, I G Evaluation of a hydrocortisone/hydroxypropyl-beta-cyclodextrin solution for ocular drug delivery Int J.Pharm 156:201–209, 1997

26 Jarho, P., Jarvinen, K., Urtti, A., Stella, V J., and Jarvinen, T Use of dextrins in ophthalmic formulations of dipivefrin Int J Pharm 153:225–233,1997

cyclo-27 Siefert, B., and Keipert, S Influence of alpha-cyclodextrin and lated b-cyclodextrin derivatives on the in vitro corneal uptake and permeation

hydroxyalky-of aqueous pilocarpine-HCl solutions J Pharm Sci 86:716–720, 1997

28 Michniak, B B., Player, M R., Godwin, D A., Lockhart, C C., and Sowell,

J W In vitro evaluation of azone analogs as dermal penetration enhancers.Part 5 Miscellaneous compounds Int J Pharm 161:169–178, 1998

29 Valenta, C., and Wedenig, S Effects of penetration enhancer on the in vitropercutaneous absorption of progesterone J Pharm Pharmacol 49:955–959,1997

30 Chatterjee, D J., Li, W Y., and Koda, R T Effect of vehicles and tion enhancers on the in vitro and in vivo percutaneous absorption of metho-trexate and edatrexate through hairless mouse skin Pharm Res 14:1058–

penetra-1065, 1997

31 Tang-Liu, D D-S., Richman, J B., Weinkam, R J., and Takruri, H Effects

of four penetration enhancers on corneal permeability of drugs in vitro J.Pharm Sci 83:85–90, 1994

32 Tang-Liu, D D.-S., and Burke, J The effect of Azone1on ocular levobunololabsorption: Calculating the area under the curve and its standard error usingtissue sampling compartments Pharm Res 5:238–241, 1988

33 Ismail, I M., Chen, C.-C., Richman, J B., Andersen, J S., and Tang-Liu, D.D.-S Comparison of azone and hexamethylene lauramide in toxicologicaleffects and penetration enhancement of cimetidine in rabbit eyes Pharm.Res 9:817–821, 1992

34 Chiou, G C Y., Chang, C Y., and Chang, M S Systemic delivery of insulinthrough eyes to lower the glucose concentration J Ocular Pharmacol 5:81–

91, 1989

35 Pillion, D J., Atchison, J A., Stott, J., McCracken, D L., Gargiulo, C., andMeezan, E Efficacy of insulin eyedrops J Ocular Pharmacol 10:461–470,1994

36 Pillion, D J., McCracken, D L., Yang, M., and Atchison, J A Glucagonadministration to the rat via eye drops J Ocular Pharmacol 4:349–358, 1992

37 Pillion, D J., Amsden, J A., Kensil, C R., and Recchia, J Structure-functionrelationship among Quillaja saponins serving as excipients for nasal and ocu-lar delivery of insulin J Pharm Sci 85:518–524, 1996.

38 Morgan, R V Delivery of systemic regular insulin via the ocular route in cats

42 Leone-Bay, A., McInnes, C., Wang, N., DeMorin, F., Achan, D., Lercara, C.,Sarubbi, D., Haas, S., Press, J., Barantsevich, E., O’Broin, B., Milstein, S.,and Patron, D Microsphere formation in a series of derivatized a-aminoacids: Properties, molecular modeling, and oral delivery of salmon calcitonin

J Med Chem 38:4257–4262, 1995

43 Leone-Bay, A., Santiago, N., Achan, D., Chaudhary, K., DeMorin, F.,Falzarano, L., Haas, S., Kalbag, D., Leipold, H., Lercara, C., O’Toole, D.,Rivera, T., Rosado, C., Sarubbi, D., Vuocolo, E., Wang, N., Milstein, S., andBaughman, R A N-acylated a-amino acids as novel oral delivery agents forproteins J Med Chem 38:4263–4269, 1995

44 Brayden D., Creed E., O’Connell A., Leipold, H., Lercara C., Agarwal R.,and Leone-Bay A Heparin absorption across the intestine: Effects of sodiumN-[8(2-hydroxybenzoyl)amino]caprylate in Caco-2 monolayers and in rat insitu intestinal instillations Pharm Res 14:1772–1779, 1997

45 Leone-Bay, A., Ho, K., Agarwal, R., Baughman, R A., Chaudhary, K.,DeMorin, F., Genoble, L., McInnes, C., Lercara, C., Milstein, S., O’Toole,D., Sarubbi, D., Variano, B., and Paton, D R 4-[4-[(2-Hydroxybenzoyl)amino]phenyl]butyric acid as a novel oral delivery agentfor recombinant human growth hormone J Med Chem 39:2571–2578, 1996

46 Robinson, J R., unpublished data, 1998

47 Yen, W.-C., and Lee, V H L Paracellular transport of a proteolytically labilepentapeptide across the colonic and other intestinal segments of the albinorabbit: Implications for peptide drug design J Contr Rel 28:97–109, 1994

48 Yen, W.-C., and Lee, V H L Penetration enhancement effect of Pz-peptide, aparacellularly transported peptide, in rabbit intestinal segments and Caco-2cell monolayers J Contr Rel 36:25–37, 1995

49 Chung, Y B., Han, K., Nishiura, A., and Lee, V H L Ocular absorption ofPz-peptide and its effect on the ocular and systemic pharmacokinetics oftopically applied drugs in the rabbit Pharm Res 15:1882–1887, 1998

50 Yen, W.-C., and Lee, V H L Role of Na+in the asymmetric paracellulartransport of 4-phenylazobenzyloxycarbonyl-L-Pro-L-Leu-Gly-L-Pro-D-Argacross rabbit colonic segments and Caco-2 cell monolayers J Pharmacol.Exp Ther J Contr Rel 36:25–37, 275:114–119, 1995

51 Saha, P., Yang, J., and Lee, V H L Existence of a P-glycoprotein drug effluxpump in cultured rabbit conjunctival epithelial cells Invest Ophthalmol Vis.Sci 39:1221–1226, 1998.

52 Zimmer, A., and Kreuter, J Microcapsules and nanoparticles used in oculardelivery systems Adv Drug Delivery Rev 16:61–73, 1995

53 Calvo, P., Alonson, M J., Vila-Jato, J L., and Robinson, J R Improvedocular bioavailability of indomethacin by novel ocular drug carriers J.Pharm Pharmacol 48:1147–1152, 1996

54 Yang, X., and Robinson, J R Bioadhesion in mucosal drug delivery In:Biomaterials for Drug Delivery (Okano, T., ed.), Elsevier, Amsterdam, 1998

55 Morimoto, K., Iwamoto, T., and Morisaka, K Possible mechanisms for theenhancement of rectal absorption of hydrophilic drugs and polypeptides byaqueous polyacrylic acid gel J Pharmacobio-Dyn 10:85–91, 1987

56 Kriwet, B., and Kissel, T Interactions between bioadhesive poly(acrylic acid)and calcium ions Int J Pharm 127:135–145, 1996

57 Lußen, H L., Bohner, V., Pe’rard, D., Langguth, P., Verhoef, J C., de Boer,

A G., Merkle, H P., and Junginger, H E Mucoadhesive polymers in peroralpeptide drug delivery: V Effect of poly(acrylates) on the enzymatic degrada-tion of peptide drugs by intestinal brush border membrane vesicles Int J.Pharm 141:39, 1996

58 Illum, L Chitosan and its use as a pharmaceutical excipient Pharm Res.15:1326–1331, 1998

59 Yoshiaki, K., Yamamoto, H., Takeuchi, H., and Kuno, Y Mucoadhesive lactide/glycolide copolymer nanospheres coated with chitosan to improve oraldelivery elcatonin Pharm Dev Tech 5:77–85, 2000

dl-60 Felt, O., Furrer, P., Mayer, J M., Plazonnet, B., Buri, P., and Gurny, R.Topical use of chitosan in ophthalmology: tolerance assessment and evalua-tion of precorneal retention Int J Pharm 180:185–193, 1999

61 Dodane, V., Khan, A M., and Merwin, J R Effect of chitosan on epithelialpermeability and structure Int J Pharm 182:21–32, 1999

62 Bernkop-Schnurch, A., Paikl, C., and Pasta, M Novel bioadhesive EDTA conjugate protects leucine enkephalin from degradation by aminopep-tidase N Pharm Res 14:917–922, 1997

chitosan-63 Bernkop-Schnurch, A., and Pasta, M Intestinal peptide and protein delivery:Novel bioadhesive drug-carrier matrix shielding from enzymatic attack J.Pharm Sci 87:430–434, 1998

Corneal Collagen Shields for

Ocular Drug Delivery

Shiro Higaki, Marvin E Myles, Jeannette M Loutsch, and

James M Hill

LSU Eye and Vision Center of Excellence, Louisiana State University

Health Science Center, New Orleans, Louisiana, U.S.A

A Soft Contact Lens for Ocular Drug Delivery

Bandage soft contact lenses made of hydrophilic polymers are widely used

to protect eyes with various problems, including recurrent corneal erosionsand epithelial defects after corneal transplantation or refractive surgery.Although these bandage soft contact lenses may enhance healing whileallowing the eye to remain open, they can be inserted and removed only

in the ophthalmologist’s office Additionally, soft contact lenses may harborpathogens, which can cause ocular infection

The idea of using bandage soft contact lenses to deliver drugs to thecornea was proposed as far back as 1971 by Kaufman (1) In this procedure,the hydrophilic lens was placed on the cornea and the drug was administeredtopically onto the surface of the lens The contact lens was thought to act as

a carrier vehicle, binding the drug and releasing it slowly, thereby increasingretention of the therapeutic agent in the tear film and at the corneal surface.However, Busin and Spitznas (2) and Matoba and McCulley (3) showedthat hydrogen contact lenses hydrated with drug are nearly devoid of drugafter only 1 or 2 hours on the cornea These soft contact lenses, therefore,are not the ideal approach for sustained, continuous ocular drug delivery

B Collagen Used to Enhance Drug Delivery

Bloomfield et al (4) were the first to suggest that collagen might provide asuitable carrier for sustained ocular drug delivery They showed that wafer-shaped collagen inserts impregnated with gentamicin produced higher levels

of drug in the tear film and tissue in the rabbit eye compared to drops,ointment, or subconjunctival injection

Fyodorov et al (5) suggested substituting collagen for hydrophilicpolymer in a contact lens shape His purpose, however, was not drug deliv-ery but the creation of a temporary protective device to enhance healing ofthe cornea In the mid-1980s, Fyodorov and colleagues (5,6) introduced thecollagen shield for use as a bandage lens and showed that the shieldsenhance corneal epithelial healing after radial keratotomy and other ante-rior segment surgical procedures

Numerous vision researchers saw the collagen shield as an extension ofand improvement on Bloomfield’s collagen inserts (4)—a potential newvehicle for the sustained administration of drugs to the cornea Over thenext several years, various drugs were incorporated into the collagen shieldmatrix during manufacture, absorbed into the shields during rehydration,and/or applied in topical drops directly onto shields in situ Studies inanimal models (described below) showed that as the drug dissolved in theshield, it was released gradually into the tear film, resulting in increasedcontact time with the cornea and increased penetration into both the corneaand the aqueous humor Clinical studies demonstrated that the collagenshield is easy to use in the ophthalmologist’s office, prevents delay in begin-ning therapy, and maintains therapeutic concentrations of drug in the eyewithout the need for frequent topical instillation of drops

PROPERTIES

A Properties of Collagen

The safety of collagen for human use is evidenced by its diverse uses andbiomedical applications Collagen is a common constituent in soaps, sham-poos, facial creams, body lotions, and food-grade gelatin In medicine, col-lagen has been used in cardiovascular surgery, plastic surgery, orthopedics,urology, neurosurgery, and ophthalmology The major medical application

of collagen is catgut suture, which is derived from intestinal collagen (7).Twenty-five percent of the total body protein in mammals is collagen; it isthe major protein of connective tissue, cartilage, and bone The secondaryand tertiary structures of human, porcine, and bovine collagen are very

similar, making it possible to use collagen derived from animal sources inhumans Biologically, collagen is suggested to promote wound healing (7).Nearly all studies on collagen have shown very low or no immunogenicity(7) Of the 10 collagen types that have been characterized, types 1, 3, and 5are the most desirable for biomedical applications because of their highbiocompatibility and low immunogenicity.

A collagen molecule consists of three polypeptide chains, called chains, which form a helix connected by interchain hydrogen bonds Thisdomain of collagen, called tropocollagen, forms a rodlike unit, 2600–26,000A˚ in length and 15 A˚ in diameter The molecular weight of the tropocollagen

a-is 300,000 daltons Collagen has a charactera-istic amino acid sequence: cine appears in approximately every third position Proline and hydroxypro-line make up approximately 25% of the total amino acid content Thehydroxyproline residues are from interchain, noncovalent cross-linkages.Newly synthesized collagen contains only a few cross-linked tropocollagenfibers However, with increased age, there is an increase in the percentage ofcross-linking (7)

gly-In the manufacture of corneal collagen shields, the ability to controlthe amount of cross-linking in the collagen subunits by exposure to ultra-violet (UV) light is an important physicochemical property, because theamount of cross-linking is related to the dissolution time of the shield onthe cornea Kuwano et al (8) investigated the effect of collagen cross-linking

on ofloxacin bioavailability In this study, the collagen shields were notimpregnated with drug, but drops were instilled after application of thecollagen shield They found the dissolution times for the cross-linked col-lagen shield were longer than those of the non–cross-linked type, therebyprolonging drug delivery times They concluded that cross-linked collagenshields might be useful ocular drug delivery devices because they can allowdrug concentrations to achieve high levels in the cornea and aqueous humor

B Properties of Commercial Corneal Collagen Shields

The collagen shield was designed to be a disposable, short-term therapeuticbandage lens for the cornea It conforms to the shape of the eye, protects thecorneal surface, and provides lubrication as it dissolves Unlike the hydro-philic plastic bandage lenses, the collagen shield offers no refractive benefit;

in fact, because it is not optically clear, it reduces visual acuity to the 20/80–20/200 range Also, the collagen shield causes some discomfort

Bio-Cor (Bausch & Lomb Surgical, Inc., Claremont, CA) was the firstcommercially available shield that was introduced in 1986 The diameter,base curve, oxygen permeability, thickness, water content, and other physi-cochemical characteristics of Bio-Cor collagen shield have been described

elsewhere (9–11) Bausch & Lomb Surgical, Inc is selling only SurgiLensnow The shields are derived from bovine collagen and are 14.5 mm indiameter Dissolution time, determined by UV irradiation during manufac-ture, is about 12 hours (Table 1) The shields are sterilized by gamma-irradiation, then dehydrated and individually packaged for storage andshipping (12) Alcon Laboratories, Inc (Fort Worth, TX) is sellingProShield The rapid dissolution as well as 12-, 24-, and 72-hour shieldsare available The shields have a diameter of 14 mm and a compoundbase curve that is approximately 9 mm when hydrated The water content

is approximately 75% (Table 1)

EXPERIMENTAL STUDIES

A variety of studies have described the pharmacokinetics of ocular delivery

of dyes and drugs by collagen shields as well as the use of the shields in thechemotherapy of various disorders These studies are reviewed below andsummarized inTables 2and3 (13,14)

A Fluorescein, a Water-Soluble Dye

To determine the ocular penetration of water-soluble compounds delivered

by collagen shields, Reidy et al (15) applied shields hydrated in a solution of

Table 1 Comparison of Characteristics of Two Collagen

Shields

Brand name ProShielda SurgiLensb

Origin of collagen Porcine sclera Bovine corium

Dissolution time (h) Rapid dissolution,

Data obtained from Alcon Surgical (Fort Worth, TX) Allergan, Inc.

(Irvine, California) and Oasis have a similar product called

KeraShield and SoftShield, respectively.

b Data obtained from Bausch & Lomb Surgical, Inc (Claremont, CA).

sodium fluorescein to normal eyes of volunteers and measured the cence in the anterior chamber by fluorophotometry The shields deliveredsignificantly larger amounts of dye to the aqueous humor at 2 and 4 hourscompared with drops of the same concentration instilled every 30 minutesover 4 hours, as well as in comparison with daily wear soft contact lensespresoaked in 0.01% fluorescein The collagen shields did not induce anydamage to the corneal epithelium over a 2-hour period These resultsdemonstrate that the collagen shield is superior to topical drops and somesoft contact lenses in delivering fluorescein to the cornea and aqueoushumor The collagen shields might also successfully deliver other water-soluble compounds, such as antibiotics, to the eye in amounts comparable

fluores-to or greater than the amounts delivered by drops over the same period oftime

B Antibacterial Agents

Ideally, chemotherapy for bacterial keratitis would delivery antibioticsrapidly to both the cornea and aqueous humor, produce concentrations ofantibiotic significantly above the minimum inhibitory concentration (MIC)

or minimum bactericidal concentration (MBC) of ocular pathogens, andsustain this high concentration for many hours However, there are numer-ous problems associated with achieving this ideal, and numerous approacheshave been taken to solve these problems (13,16,17)

Various investigators have examined the utility of the collagen shieldfor the delivery of antibiotics to the cornea and aqueous humor In one ofthe earliest pharmacokinetic studies, Unterman et al (18) assessed the phar-macokinetics of tobramycin delivered to rabbit eyes by means of collagenshields hydrated in solutions of either 40 or 200 mg/mL of tobramycin.Tobramycin concentrations in the cornea and aqueous humor were deter-mined at 2, 4, and 8 hours after application No toxicity was observed withshields hydrated in the 40 mg/mL solution at any time Eight hours afterapplication, the corneas with shields hydrated in the 200 mg/mL solution oftobramycin had some epithelial defects At all times and with either hydra-tion solution, the concentration of tobramycin in the cornea and aqueoushumor exceeded the MIC for most aminoglycoside-sensitive strains ofPseudomonas

O’Brien et al (19) compared collagen shields with soft contact lenses inpharmacokinetic studies of the ocular penetration of tobramycin Threegroups were compared: (a) eyes with collagen shields rehydrated in 3 mg/

mL of tobramycin, (b) eyes with therapeutic soft contact lenses, and (c) eyeswith neither lenses nor shields Topical tobramycin (3 mg/mL) was applied

to all eyes every 5 minutes for a total of six doses Aqueous humor samples