drug release rate in vivo, rate of drug clearance via tear turnover, drug clearance from the lacrimal fluid to the cornea, drug clearance from the lacrimal fluid to the conjunctiva, drug permeability in the conjunctiva, conjunctival surface area, corneal permeability of the drug, corneal surface area.
In equation 2 clearances across the membranes are calculated as the products of membrane permeabilities (P) and their surface areas (5).
If the dosage form flows partly or completely from the eye after administration, as viscous vehicles, gelling systems, nanoparticulates and liposomes do, it is very important that the drug release rate and the rate of dosage form drainage from the eye match each other (Fig. 1). Too great a drainage rate of the controlled release dosage form from the conjunctival sac may result in poor ocular bioavailability because the drug is not adequately released from the system to the lacrimal fluid before the dosage form is removed. For example, typical liposomes and nanoparticles have half-lives of 3-5 min and 2min, respectively, on the ocular surface (Fitzgerald et al., 1987a,b). Sometimes drug is not adequately released from liposomes in the tear fluid and, consequently, ocular absorption of liposomal epinephrine (Stratford etal., 1983) and dihydrostreptomycin (Singh and Mezei, 1984) is less than from conventional eyedrops. Release rates of drugs or fluorescent probes from liposomes in vitro in buffer solutions are in the order of a few per cent per hour (Fitzgerald et al., 1987a), but in vivo release rates may be much higher because the tear fluid increases the rate of drug release from liposomes as much as does plasma (Barber and Shek, 1986). The half-life of carboxyfluorescein release from multi-lamellar vesicles decreases from 6.96 ± 1.03 h to 0.43 ± 0.05 h when the percentage of tear fluid in the release medium increases from 0% to 100%. The mechanism for the increased rate of drug release in lacrimal fluid is not known. Also, in the where dQ/àt = Chi = Clco = Clci =
case of colloid size particles, endocytosis by the conjunctival cells (Latkovic and Nilsson, 1979) may increase the particle retention and drug absorption via conjunctiva and sclera into the eye (Ahmed and Patton, 1986).
Several vehicles have been described that are capable of prolonging drug retention and drug activity in the eye (Shell, 1984; Lee and Robinson, 1986). Viscous, mucoadhesive gels or 'after application' gelling vehicles increase ocular drug absorption by prolonging ocular contact (Lee and Robinson, 1986) but, strictly speaking, these dosage forms are not controlled release systems. They increase both ocular peak drug concentrations and bioavailability, because the drug release from the vehicle is rapid compared with the rate of drug penetration into the eye and, thus, drug release is not the rate-controlling factor in ocular drug delivery (Maurice and Mishima, 1984). This kind of profile has been called prolonged pulse entry of drug into the eye (Salminen etal., 1983).
In contrast to nanoparticles and liposomes, drug release from these systems is faster than their precorneal drainage and drug release is not crucial because the release is rapid. Retention of the vehicle or dosage form on the ocular surface can be misleading. For example, a mucoadhesive hyaluronic acid vehicle shows excellent retention for an eyedrop on the ocular surface (Gurny etal., 1987). Based on gammascintigraphic data, only about 20% of the vehicle had been eliminated from the eye after 10 min whereas the half-life of a saline eyedrop was about 30 s. Despite this order of magnitude difference in the rate of vehicle removal from the eye, ocular biological activity of pilocarpine measured as an area under the miosis versus time curve was improved only 25% and 61% in man and rabbit, respectively, with the hyaluronic acid vehicle. Careful interpretation of the gammascintigraphic data is needed, because the label (usually ["mTc]-DTPA complex) does not penetrate the ocular tissues. In contrast, pilocarpine readily absorbs (about 24% min ~1 at a volume of 30 nl) across the conjunctiva to the systemic blood circulation (Fig. 1) (Urtti etal., 1985). Pilocarpine concentration in the tear fluid may decrease much more rapidly than the measured 99111 Tc or hyaluronic acid concentration resulting in only modest increases in pilocarpine absorption.
The rate of drug release from some polymeric controlled release systems may be lower in the lacrimal fluid than in in vitro release tests; this may be for several reasons. For example, the mixing conditions in the conjunctival sac are mild and tear volume is very small (7 /¿l) (Maurice and Mishima, 1984). Monoesters of polyvinyl methyl ether/maleic anhydride) release timolol at lower rates in vivo in rabbit eyes than in vitro in phosphate buffer (Finne and Urtti, 1992). Because these polymers are polycarboxylic acids their dissolution is dependent on surface pH and transport of hydrogen ions across the static diffusion layer on the polymer surface (Heller etal., 1978).
Consequently, low buffering capacity of the tear fluid and poor mixing result in conditions where both surface pH may decrease and the thickness of the diffusion layer may increase. Consequently, the rate of polymer dissolution and drug release are slowed down in vivo (Finne and Urtti, 1992). This factor may affect the drug release from other ionizable and bioerodible polymers in the tear fluid as well.
In contrast to erodible polymer matrices, similar in vitro and in vivo release rates have been demonstrated for pilocarpine from Ocusert (Urquhart, 1980) and for timolol from silicone reservoir devices (Urtti etal., 1988; Urtti etal., 1994). These polymers neither dissolve nor have interactions with the lacrimal fluid. Here there is no interphase between dissolving solid polymer and already dissolved polymer in the diffusion layer. On the surface of non-erodible reservoir devices there is a diffusion layer with a concentration gradient of the drug but, in most cases, drug concentration in the lacrimal fluid on the polymer surface is substantially lower than its solubility. This hydro-dynamic diffusion layer does not limit drug release of timolol from silicone reservoir devices in vivo in lacrimal fluid (Urtti etal., 1988, 1994).
In principle, drug release from diffusion controlled delivery systems is determined by the thermodynamic activity gradient of the drug between outer polymer surface and lacrimal fluid. If there is a drug concentration buildup in the receptor medium, drug release rate should decrease because the concentration gradient is decreased. However, owing to the typically high drug clearances from the lacrimal fluid (5-10^1 min~') (Lee and Robinson, 1979; Urtti etal., 1990; Keister etal., 1991) mainly via conjunctival drug absorption, no significant concentration buildup is expected to take place at drug release rates less than 1 mg h-1 (Urtti, 1991). These calculations are supported by the data on non-erodible timolol inserts of silicone that released timolol at the rates of 4-7 /igh-1. The release rates in vitro and in vivo were similar both in rabbit (Urtti etal., 1988) and human (Urtti etal., 1994) eyes. This contrasts with transdermal drug delivery in which poor transdermal drug permeability and subsequent concentration buildup in the occluded sweat may suppress the release rate of the drug from a transdermal patch (Urtti, 1991).
After application of solid inserts, drug concentration in the lacrimal fluid of rabbits is not homogeneously distributed (Urtti etal., 1988). Mixing in the preocular fluid of rabbits is poor owing to the infrequent blinking (four times per minute) and slow tear turnover. This was demonstrated in a study where a silicone device releasing timolol was placed either in the superior or inferior conjunctival sac in pigmented rabbits (Urtti etal., 1988). Application to the superior cul-de-sac resulted in greater corneal timolol absorption compared with the device placement in the inferior conjunctival sac. From the inferior conjunctival sac, timolol absorption took place mainly via the inferior part of the conjunctiva and sclera and thus timolol concentrations in the lower parts of each ocular tissue (cornea, conjunctiva, sclera, iris-ciliary body) were much higher than in the upper parts of the tissues. Poor mixing in the rabbit tear film has been demonstrated also by Lutosky and Maurice (1986). Drug distribution after application of an ocular insert to the human eye may be more even because the blinking frequency is several times per minute (Saettone etal., 1982).
Was this article helpful?