A schematic and simplified presentation of ocular pharmacokinetics related to topical ocular administration is shown in Fig. 1.
DRUG IN DOSAGE FORM
RELEASED DRUG IN TEAR FLUID
REMOVAL BY DRAINAGE, BLINKING, INDUCED LACRIMATION, AND TEAR TURNOVER
■ LACRIMAL DRAINAGE SYSTEM
fig. 1 Schematic presentation of ocular pharmacokinetics related to topical ocular drug administration.
When eyedrops are administered to the ocular surface the solution mixes with the tear fluid. After administration the extra solution volume rapidly flows from the eye (Chrai etal., 1973, 1974) through the puncta to the lacrimal drainage system and then to the nose, pharynx and gastrointestinal tract (Hurwitz etal., 1975; Chavis etal., 1978). The drainage of the instilled solution is rapid: the drainage rate constant increases with instilled volume - 0.31 min"'and 0.82min"1 for eyedrops of 5¡A and 50/il, respectively, in rabbits (Chrai etal., 1973). Furthermore, the rate of solution drainage decreases with elevated solution viscosity (Chrai and Robinson, 1974) and mucoadhesiveness (Gurny etal., 1987). Drainage rate of ophthalmic solutions is about three times faster in humans than in rabbits (Zaki etal., 1986). This is partly explained by the higher blinking frequency in humans than in rabbits (Chavis etal., 1978; Saettone etal., 1982).
Another important route of drug loss from the lacrimal fluid is drug absorption to the conjunctiva (Lee and Robinson, 1979; Thombre and Himmelstein, 1984). The conjunctiva of the eye is lined by stratified columnar epithelium two to seven cell layers thick, depending on the conjunctival region (Junqueira etal., 1986). Conjunctival permeability is fairly high so that most drugs diffuse across the conjunctiva easily (Wang etal., 1991). Typical values for conjunctival drug permeabilities are in the order of 1-5 x 10~5cm s-1 (Ahmed etal., 1987; Wang etal., 1991). Consequently, nearly 50% of topically instilled" pilocarpine is absorbed within a few minutes through conjunctiva to systemic circulation in rabbits (Urtti etal., 1985).
Induced lacrimation may increase the rate of solution drainage from the preocular area for several reasons. Often eyedrop solutions are buffered to a low pH for reasons of stability. Examples of this are pilocarpine, atropine and epinephrine (Dolder and Skinner, 1983). Acidic pH may, however, cause discomfort and thereby increase lacrimation and rate of drug removal from the corneal surface (Conrad etal., 1978). Also, if the osmotic pressure of instilled eyedrops deviates from isotonicity, the rate of instilled solution drainage is increased (Conrad etal., 1978). Topical anaesthesia blocks the induction of lacrimation for the most part and results in a reduced precorneal drainage rate of the instilled solution and in increased ocular pilocarpine absorption in rabbits (Patton and Robinson, 1975). The basal tear turnover has only a minor role in the drug removal from the preocular area (Lee and Robinson, 1979).
At the time of eyedrop instillation there is a concentration gradient between the tear fluid and neighbouring corneal and conjunctival tissues. Consequently, the difference in thermodynamic activity drives drug to these tissues. Thus, for example, in the case of pilocarpine most of the corneal absorption takes place within 3 min after instillation of an eyedrop in rabbits (Chrai etal., 1974). The period of drug absorption is short because the activity gradient decreases rapidly owing to precorneal solution drainage and conjunctival systemic absorption (Chrai etal., 1973; Lee and Robinson, 1979). The initial rate of decrease in lacrimal drug concentration (e.g. pilocarpine) is typically 0.4-1.0min-1 (Chrai etal., 1973; Lee and Robinson, 1979; Urtti and Salminen, 1985). Thereafter, the drug concentration in the tear fluid decreases much more slowly because pseudo-equilibrium between tear fluid and surrounding tissues has been established (Urtti etal., 1990). In the terminal phase the elimination rate of timolol from the lacrimal fluid of rabbits is even slower (0.003 min-1) than the rate of normal tear turnover (0.07min-1) suggesting considerable back-diffusion from the cornea and conjunctiva to the tear fluid (Urtti etal., 1990). It should be remembered that despite the sustained drug concentrations in the lacrimal fluid after eyedrop administration no more drug absorption takes place during this kinetic phase.
The cornea is the main route of drug absorption from the tear fluid to the inner eye in most cases (Doane etal., 1978). Examples of drugs absorbed through this route are timolol (Ahmed and Patton, 1985), pilocarpine (Doane etal., 1978) and hydrocortisone (Doane etal., 1978). For most drugs the main corneal penetration barrier lies in the epithelium (Sieg and Robinson, 1976). The importance of different epithelial cell layers as barriers is dependent on the lipophilicity of compounds (Shih and Lee, 1989). In the case of hydrophilic atenolol the entire epithelium behaves as the penetration-limiting barrier, whereas for timolol and levobunolol, compounds with intermediate lipophilicity, only the most superficial cell layers considerably limit drug permeability (Shih and Lee, 1990). These outermost cellular layers of corneal epithelium are interconnected by tight junctions which limit especially the penetration of hydrophilic molecules (Grass and Robinson, 1988; Wang etal., 1991). In the epithelium wing cells and basal cells are less tightly interconnected and allow intercellular penetration of macromolecules such as horse radish peroxidase (Tonjum, 1974) and fluorescein isothiocyanate (FITC)-labelled poly(L-lysine) (Rojanasakul etal., 1990).
Corneal stroma is a loosely arranged hydrophilic layer with no continuous cellular structure. Consequently, drug diffusion in the stroma is rapid and values are in the typical range not dependent on lipophilicity and molecular weight (Maurice and Mishima, 1984).
Many ophthalmic drugs show considerable lipophilicity that makes trans-cellular drug penetration possible. Because of the differences in transcel-lular permeabilities, increasing lipophilicity has been shown to enhance corneal permeability so that maximal permeability of /3-blockers was observed at log (octanol/water) partition coefficients of 2-3 (Huang etal., 1983). Similar findings were observed with steroids (Schoenwald and Ward, 1978). Data sets of more heterogeneous groups of compounds revealed substantial deviations from the simple partitioning effects caused by different molecular sizes (Grass and Robinson, 1988) and charges (Liaw etal., 1991). As expected, increased molecular weight decreases the corneal permeability (Grass and Robinson, 1988; Maurice and Mishima, 1984). At neutral pH the rabbit cornea behaves as if it were negatively charged. Consequently, positively charged L-lysine had a 10 times higher permeability than the negatively charged L-glutamic acid (Liaw etal., 1991).
In the range of very lipophilic compounds (logP > 3) corneal permeability cannot be further improved by increasing the lipophilicity; steady (Wang etal., 1991) or decreasing (Schoenwald and Ward, 1978; Huang etal., 1983) permeabilities at higher lipophilicities have been observed. This phenomenon is explained by impaired drug desorption from the lipophilic parts of the corneal epithelium to the hydrophilic stroma (Huang etal., 1983). Therefore, the corneal stroma appears to be the penetration rate limiting barrier in these cases. It should be remembered, however, that ocular bioavailability is not directly related to corneal permeability. In addition to the corneal permeability, the ocular bioavailability is affected by the precorneal drainage factors and by the ratio of permeabilities between the cornea and conjunctiva (Fig. 1) (Wang etal., 1991).
Corneal permeability of drugs can be improved by optimization of the formulation pH so that the fraction of the unionized drug is increased (Francouer etal., 1983; Ashton etal., 1991). Another more demanding approach is to prepare a prodrug derivative that has improved corneal absorption characteristics and releases the active parent compound through enzymatic and/or chemical hydrolysis in the eye (Lee and Li, 1989). These chemical delivery systems also provide potential to control drug input rate and prolong drug action in the eye.
The conjunctiva covers most of the ocular surface and has greater permeability than the cornea, especially for hydrophilic compounds (Wang etal., 1991). Consequently, depending on dosing conditions and dosage form, 5-30 times more timolol (Chang and Lee, 1987; Urtti etal., 1990) and pilocarpine (Lee and Robinson, 1979; Thombre and Himmelstein,
1984; Urtti etal., 1985) is absorbed through the conjunctiva to the systemic circulation than transcorneally into the eye. Conjunctival clearance of the drug from the lacrimal fluid is determined as CL = P x S, where P is the conjunctival permeability (typically in the order of 10 cm s_1) (Wang etal., 1991) and S is the conjunctival surface area (18 cm2) (Watsky etal., 1988). Furthermore drug flux through the conjunctiva at steady-state is determined in clearance terms as J = CL x dC, where dC is the concentration gradient of drug between the lacrimal fluid and conjunctiva.
Part of the drug may absorb via conjunctiva to the sclera and thereafter to the ciliary body (Ahmed and Patton, 1985). Drug that is absorbed by this route does not usually gain access to the aqueous humour. Conjunctival and scleral penetration is more favourable than the corneal route, especially for hydrophilic and large molecules like inulin that have poor corneal permeability (Ahmed and Patton, 1985; Ahmed etal., 1987). In addition, conjunctiva has been shown to be the preferable route for ocular absorption of the hydrophilic a2 agonist p-aminoclonidine (Chien etal., 1990).
After absorption into the aqueous humour a drug may distribute to the surrounding tissues: iris, ciliary body, and lens. A small amount of the drug may also penetrate further to the posterior chamber and vitreous humour (Maurice and Mishima, 1984).
Penetration to the iris and ciliary body takes place easily because these tissues have a porous, leaky surface (Maurice and Mishima, 1984). Consequently, drug concentrations in the iris and ciliary body of albino rabbit reflect those in the aqueous humour readily and they are considered to belong to the same pharmacokinetic compartment (Makoid and Robinson, 1979). Deviations from this relationship take place in the case of pigmented iris and ciliary body if the drug is capable of binding to melanin. For example, pilocarpine (Lee and Robinson, 1982), timolol (Salminen and Urtti, 1984), and atropine (Salazar and Patil, 1976) bind to ocular pigmentation. Typically, binding results in elevated drug concentrations in, and in slower drug elimination from, the tissue (Lee and Robinson, 1982; Salminen and Urrti, 1984). This results in prolonged action of pilocarpine (Urtti etal., 1984) and atropine (Salazar and Patil, 1976) in pigmented eyes compared with less pigmented ones. Thus, pigmentation may behave like an intraocular sustained release depot for some drugs.
The lens is less permeable than iris or ciliary body and, consequently, drug concentrations in this tissue are typically an order of magnitude lower than in the anterior uvea (Maurice and Mishima, 1984; Urtti etal., 1990). The lens decreases drug diffusion from the anterior chamber to the vitreous humour (Maurice and Mishima, 1984). Consequently, higher concentrations are achieved in the back of aphakic than phakic eyes after topical ocular drug administration. Also, convective flow of the aqueous humour from the posterior to the anterior chamber impairs drug access to the posterior eye. Thus, it is not surprising that topical ocular drug administration results in drug concentrations nearly two orders of magnitude lower in the vitreous than in the aqueous humour (Urtti etal., 1990). Consequently, it is difficult to treat disorders in the back of the eye using topical ocular drug administration, while drug treatment of the posterior eye via systemic circulation is impaired by the blood-vitreous barrier (Maurice and Mishima, 1984).
Drugs are eliminated from the aqueous humour by aqueous drainage through the trabecular meshwork. Drug clearance by this route equals the rate of elimination from the aqueous humour (Conrad and Robinson, 1977; Miller etal., 1981). In addition, many drugs, like pilocarpine and timolol, are eliminated via the blood circulation of the anterior uvea (Miller etal., 1981; Tang-Liu etal., 1984). Consequently, clearance values of pilocarpine, timolol, and flurbiprofen are more than 10^1 min-1, which is several times higher than the elimination rate from aqueous humour of the rabbit eye (Tang-Liu etal., 1984). In ocular pharmacokinetics it is difficult to distinguish the distribution phenomena from elimination because the lens and vitreous humour can form large and slowly equilibrating drug depots.
After intracameral injection and topical eyedrop application, drug concentration in the aqueous humour decreases according to biphasic kinetics with two half-lives - one for drug distribution and the other for elimination (Makoid and Robinson, 1979; Miller etal., 1981). For example, the apparent half-life for flurbiprofen distribution is 15 min and the half-life for its elimination from the aqueous humour after intracameral injection is 93min (Tang-Liu etal., 1984).
Although drug absorption from the lacrimal fluid to the cornea ceases in a few minutes, the second step in ocular drug absorption, transfer from the epithelium to the stroma, is slow and may decrease the value of the apparent half-lives of distribution and elimination (Makoid and Robinson, 1979). During the terminal elimination phase the half-life is prolonged by the back-diffusion of the drug from the tissue reservoirs such as the lens, vitreous humour and pigmented uvea (Makoid and Robinson, 1979; Miller etal., 1981).
Was this article helpful?