After eyedrop application drug concentration in the tear fluid decreases rapidly owing to efficient solution drainage and absorption processes (Fig. 1). Consequently, most of the corneal absorption takes place during the first few minutes after instillation of the eyedrop (Chrai etal., 1974).
When viscous and mucoadhesive solutions, gels, or thermosetting gels are used the initial rate of drug removal is slower and more drug may be absorbed by the cornea (Chrai and Robinson, 1974; Saettone etal., 1982; Shell, 1984; Gurny etal., 1987). The possibilities of increasing ocular absorption are, however, dependent on the pharmacokinetic properties of the drug. Very lipophilic drugs rapidly attain their partitioning equilibrium with corneal epithelium and no further increase of drug absorption is achieved by modest prolongation of their corneal contact (Grass and Robinson, 1984; Keister etal., 1991). With a solution of viscosity 90CP, improvement of the corneal drug absorption was achieved only with water-soluble drugs. Drugs with a log (octanol/pH 7.4 phosphate buffer) distribution coefficient above 1.0 did not show increased ocular absorption when administered in eyedrops with elevated viscosity. Ocular bioavailability is also dependent on the vehicle effects on conjunctival systemic absorption (Ashton etal., 1991).
Without rate control of drug release it is, however, difficult to achieve sustained drug concentrations in the eye. In theory, sustained drug levels and substantial increase in ocular bioavailability can be achieved if the precorneal drug loss rate is 0.1 h-1 or less (Maurice and Mishima, 1984). In practice, conjunctival non-productive drug absorption makes this goal impossible. For example, Chang and Lee (1987) blocked the drainage of timolol from the rabbit conjunctival sac mechanically for 2h, but this resulted in only about a threefold increase in ocular absorption and the apparent half-life of timolol elimination decreased by 15%. In addition, the half-lives of most ocular drugs in the aqueous humour are in the order of
1 h. Consequently, doubling the peak levels extends the concentration curve only by 1 h (Maurice and Mishima, 1984). However, it should be remembered that often the pharmacological activity decays more slowly than the drug concentration in the aqueous humour. This is the case particularly in pigmented eyes and may result in more substantial prolongation of the pharmacological effects through prolonged corneal contact (Urtti et al., 1984).
In conclusion, adequate retention in the conjunctival sac and control of release rate are required in order to achieve sustained drug concentrations, increased duration of activity and decreased peak drug levels in the eye. Otherwise, only limited increase in the duration of activity is achieved, while increased dose and/or improved bioavailability result in the elevation of peak drug concentrations and possible side-effects in the same proportion as absorption was increased.
For controlled release systems the fraction of the corneal absorption at steady state is determined as C/c/(C/cj + C/tf+C/c). The equation gives the maximal ocular corneal absorption that can be achieved with the controlled release system staying constantly in the conjunctival sac (Fig. 1). For example, the maximal ocular bioavailability of timolol is 11 °7o, which is approximately 2.5 times higher than the bioavailability after eyedrop administration (Urtti etal., 1990).
Consequently, it is important to minimize the conjunctival permeability relative to the corneal permeability. The corneal drug permeability is more sensitive than the conjunctival permeability to the effect of lipophilicity (Wang etal., 1991). Thus, the fraction of ocular absorption is increased with increasing lipophilicity and also when the drug is administered in a controlled release system. It is obvious that the main factor that limits the bioavailability in the case of controlled release inserts is conjunctival nonproductive drug absorption (Fig. 1). Theoretical calculations by Keister etal. (1991) demonstrate that, in the case of very lipophilic drugs, neither increased ocular contact nor decreased eyedrop volume increase the ocular bioavailability. Consequently, the ocular bioavailability of lipophilic drugs is not necessarily increased with ocular controlled release systems, even though the duration of activity and the shape of the response versus time curve is improved. The clearance ratio determines the fraction that is absorbed transcorneally, but the steady-state drug concentration in the aqueous humor is also affected by the rate of drug release (Urtti etal., 1990).
Drug concentration in the aqueous humour is not a good indicator of ocular bioavailability when the drug is absorbed via bulbar conjunctiva and sclera to the eye (Fig. 1). Ophthalmic controlled release systems are usually in contact with the conjunctiva. Conjunctival contact may result in increased drug loss to systemic circulation but conversely it may improve the ocular drug delivery via conjunctiva and sclera. This is an interesting possibility in the ocular delivery of large polar molecules like peptides. For example, selective non-corneal delivery of inulin with liposomes to the iris and ciliary body has been demonstrated by Ahmed and Patton (1986).
Collagen shields are lens shaped and they are placed on the cornea to protect and to facilitate epithelial healing. Usually the collagen is obtained from porcine sclera. Collagen shields have been studied as potential drug carriers since the late 1980s. Usually, water-soluble drugs are incorporated into the collagen shield by soaking the shield in drug solution. After hydration the water content of the shield is 65-83% (Friedberg etal., 1991). Drug is released after the shield is placed on the cornea. Water-soluble drugs diffuse rapidly from the collagen shields so that their half-life is short, e.g. tobramycin in a 'MediLens' shield has a half-life of only 5 min (Assil etal., 1992). Poorly water-soluble drugs like cyclosporin A are incorporated in the shield during preparation and are released more slowly than more water-soluble drugs (Reidy etal., 1990). Drug release can be modified also by binding the drug to the shield and by changing the degree of cross-linking of the collagen.
Kinetically, collagen shields are exceptional because they are in direct contact with the cornea. Theoretically, the corneal drug absorption relative to conjunctival systemic absorption should be maximized. Gentamicin, tobramycin, vancomycin, dexamethasone, cyclosporin A, and heparin have been administered ocularly in collagen shields (Friedberg etal., 1991). Administration by collagen shields results in increased or comparable ocular drug absorption compared with eyedrops (Friedberg etal., 1991).
There are, however, some potential problems associated with collagen shields. These include mass production problems and high price. In addition, Assil etal. (1992) have shown that collagen shields cause diffuse punctate epitheliopathy which may be one reason for improved corneal drug penetration. These factors may limit the use of collagen shields to selected cases. The most promising area is the postoperative antibiotic treatment of the eyes.
Was this article helpful?