Info

a EXT = extrapolated method; determined by extrapolating log-linear elimination phase t0 C at t0 and dividing into the intracameral dose. b SS = steady-state method; determined by maintaining a constant concentration of drug on the cornea of an anesthetized rabbit and measuring aqueous humor concentrations of drug over time during infusion and postinfusion. See Table 3 for equatiaon for VSS.

a EXT = extrapolated method; determined by extrapolating log-linear elimination phase t0 C at t0 and dividing into the intracameral dose. b SS = steady-state method; determined by maintaining a constant concentration of drug on the cornea of an anesthetized rabbit and measuring aqueous humor concentrations of drug over time during infusion and postinfusion. See Table 3 for equatiaon for VSS.

The pharmacokinetic parameter, Vd, also provides a general assessment of distribution. However, it does not differentiate between sites that are active and those that are either devoid of activity of potentially responsible for side effects. In lieu of making these measurements, it is not difficult to measure eye tissue concentrations over time, which provides a direct indication of whether or not drug distributes to the active site. In contrast to the paucity of pharmacokinetic measurements for ophthalmic drugs, tissue concentrations of drug have been widely reported for many years. With recent improvements in assay methodology, tissue profiles over time are routinely measured for aqeous humor and cornea, but less frequently for lens, iris/ciliary body, conjunctiva, choroid, lacrimal gland, and/or retina.

In the interest of therapy, it is of importance to determine the percent of the dose that reaches a particular tissue and to determine if the concentrations that are reached meet and/or exceed the minimum therapeutic

Figure 5 Average (n = 5) aqueous humor concentrations of amikacin (2 mg/mL) and chloramphenicol (5 mg/mL) following direct injection of into the anterior chamber of the rabbit eye (10 mL). (Adapted from Ref. 5.)

range. For example, in a recent study by Morlet et al. (77), ciprofloxacin, which is active against a wide variety of bacteria, was given orally in doses of 750 or 1500 mg prior to eye surgery and then assayed for levels in blood, aqueous, and vitreous humor during surgery. Population pharmacokinetics was applied to the data, and the mean half-lives for loss of ciprofloxacin were determined for aqueous and vitreous humor, which were found to be 3.5 and 5.3 hours, respectively. At steady state, the concentrations of drug in aqueous and viterous were 23 and 17% of serum levels, respectively. The authors concluded that the use of oral ciprofloxacin would be below the minimum inhibitory concentration (MIC) of 0.5 mg/mL in treating bacterial endophthalmitis or for use for perioperative prophylaxis.

It is reasonable to expect that if drug is instilled topically to the eye and no unusual tissue affinity occurs for a particular drug, the order of decrasing tissue concentrations are as follows: cornea > conjunctiva > aqueous humor > iris/ciliary body > lens, vitreous, and/or choroid/retina. Many studies (37,51,52,67,78,80,81) show that iris/ciliary body concentrations are higher than aqueous humor concentrations, even though the dose is instilled topically to the cornea. A number of reasons have been suggested to account for this phenomenon. For example, the drug may distribute extensively to iris/ciliary body, but drug may equilibrate rapidly with aqueous humor so that the elimination rate during distribution equilibrium is the same as elimination from aqueous humor. If this explanation is correct, the iris/ciliary body would have a large capacity for drug but not exhibit an unusually high binding affinity. It is also conceivable that the binding affinity as well as its capacity is very high so that the elimination rate from iris/ciliary body and aqueous are not the same but drug remains in the former tissue much longer.

This latter observation was observed by Putnam et al. (78) for amino-zolamide, a topically active carbonic anhydrase inhibitor (CAI), which showed a relatively slow elimination for drug as well as metabolite, 6-acet-amido-2-benzothiazolesulfonamide, from iris/ciliary body compared to aqueous humor following topical administration to the rabbit eye. In another example, a derivative of methazolamide, 5-acetoxyacetylimino-4-methyl-A2-1,3,4,-thiadiazoline-2-sulfonamide, is an ester prodrug that lowered intraocular pressure in albino New Zealand rabbits but was found to be active in pigmented Dutch Belt rabbits (79). From various studies it was concluded that melanin binding in the iris, and to some extent metabolism occurring only in the Dutch Belt rabbits, were valid explanations.

Alternatively, support for the possibility of high concentrations of drug in the iris-ciliary body after topical application come from the likelihood that drugs reach iris-ciliary body by scleral absorption (i.e., paracel-lular penetration) and/or uptake by vessels imbedded in the conjunctiva that deposit drug. Certain drugs, when applied topically to the eye, may be preferentially absorbed by the sclera, as opposed to the cornea, and enter the iris/ciliary body without first entering the aqueous humor. In a study by Ahmed et al. (71), corneal and scleral penetration were determined for propranolol, timolol, nadalol, penbutolol, sucrose, and inulin. The results of the study showed that the outer layer of the sclera provides much less resistance to penetrabiilty for hydrophilic drugs than the corneal epithelium. Hamalainen et al. (30) estimated that the conjunctival and scleral tissues were 15-25 times more permeable than the cornea and that conjunctival permeability was less affected by molecular size than the cornea. In addition, the total paracellular space of the conjunctiva was estimated to be 230 times greater than that in the cornea. For lipophilic drugs, such as propranolol, timolol, and penbutolol, the difference in penetrability between the tissues has been estimated to be similar. In a study by Schoenwald and Zhu (52), ketanserin and its metabolite, ketanserinol, were infused topically to anesthetized rabbits for 120 minutes. Drug was instilled either within the well and, therefore, in direct contact with the cornea, or outside the well, excluding the cornea but allowing drug to come in contact with conjunctiva

(see Fig. 3). The results showed that iris-ciliary body concentrations were three and two times higher for ketanserin and ketanserinol when applied exclusively to the sclera as opposed to the cornea (52).

In a similar study by Chien et al. (79), various anterior chamber tissue levels were measured over time for clonidine, p-aminoclonidine, and a 6-quinoxalinyl derivative of chlonidine (AGN 190342) using the infusion method. Whenever drug was maintained on conjunctiva or cornea for a period of 60 minutes, tissue concentration followed a clear trend. the order of highest to lowest concentration of drug following conjunctival contact was conjunctiva > cornea > ciliary body > aqueous humor. When drug solution was in contact with cornea only, the order was cornea > aqueous humor > ciliary body > conjunctiva. From the results of recent studies (30,31,7982), various pathways of ocular absorption have been proposed, as summarized in Figure 6.

Although many anterior and, to a lesser extent, posterior tissues have been measured for drug concentration over time following topical or systemic administration, little definitive information exists regarding the ability of tissues to accumulate drug through either partitioning or binding of drug. Of critical importance is the iris/ciliary body, which is the biophasic location for many pharmacological responses acting on the eye. Also of interest is the lens, since drug accumulation may be responsible for inducing cataract formation. Both of these tissues are easily removed and often treated as a kinetically homogeneous tissue, but drug concentrations within these tissues are likely to be quite variable in concentration since these tissues are not anatomically homogenous.

For example, iris tissue in the rabbit eye is porous and highly vascular with a large surface area in direct contact with aqueous humor; consequently, distribution equilibrium between iris and aqueous humor should occur rapidly. Also, as the iris becomes darker, the capacity for pigmented iris to bind to catecholamines increases. This was also observed for carbonic anhy-drase inhibitors (CAIs) (82). Dark-eyed individuals have a delayed onset and a reduced but prolonged response to catecholamines, presumably due to the reservoir effect that the pigmented iris has on ocular disposition of catecho-lamines. Latanoprost, an isopropyl ester of prostglandin F2a, can darken the iris in susceptible individuals due to increased pigmentation of the anterior surface of the iris (83). In contrast to iris tissue, the ciliary body epithelial layer of the pars plicata, containing a high concentration of carbonic anhy-drase and presumed to be the active site for CAIs, is not easily reached via topical instillation (84), requiring a lipophilic drug substance to readily penetrate cellular membranes. Although the separation of these tissues from one another would provide useful information, it is usually not done because of the difficulty in separating one tissue entirely from the other.

Figure 6 Probable ocular penetration routes for topically applied drugs: (1) trans-comeal pathways; (2) noncorneal pathways; (3) systemic pathways.

The entire lens is also easy to remove during kinetic studies; nevertheless, the lens should not be treated as a kinetically homogeneous tissue. Maurice and Mishima (85) reported that fluorescein, a water-soluble dye, spreads laterally and rapidly in the outer layers of the lens but does not diffuse readily into the lens nucleus. The initial barrier for entry into the lens is the epithelium, which is a single layer of cells lying just below the anterior lens capsule. Internal to the epithelial layer are densely packed lens fibers and the nucleus. Because it consists of hard, condensed cellular material, the nucleus possesses high tortuosity and low porosity. With age, old fibers are not disposed of but become compressed centrally to form a larger, less elastic nucleus. Consequently, drugs are not expected to readily penetrate the lens interior. Ahmed et al. (86) came to this conclusion from a study of the diffusion of timolol through the lens fibers of rabbit lens. A pharmacokinetic model for drug entering aqueous humor and distributing into the lens included the capsule and its epithelium as a compartment adjacent to aqueous, but excluded the lens interior since drug did not reach significant levels in this region. It was concluded that unless a steady state of drug is present and maintained through repetitive dosing, significant accumulation would not occur. Drug elimination from the eye is too rapid to allow for significant concentrations to accumulate into the interior of the lens from a few doses.

3. Elimination

Drug is eliminated from the anterior chamber, at least in part, by aqueous humor turnover, which in the rabbit eye is 1.5% of the volume of the anterior chamber per minute or expressed as a half-life, 46.2 minutes (87). Although clearance is ofgreatest interest for drugs used systemically, half-life representing loss from aqueous humor is the most common pharmacokinetic parameter measured following topical administration to the eye. Table 5 lists half-lives for drugs of ophthalmic interest. Surprisingly, nearly all the drugs fall within a range of 0.6-3 hours, which is less than when these same drugs are studied systemically. Either there are few tissue-binding sites in the eye to allow for drugs to resist clearance, or the pathways by which drugs are eliminated are very efficient. The latter may be the most likely explanation, since aqueous humor volume is relatively large compared to ocular tissue volume.

Elimination can also be expressed as a clearance, and if the anterior chamber of the rabbit contains about 0.311 mL, an average aqueous humor clearance due to bulk flow becomes 4.67 mL/min based upon Eq. (4). Ocular clearances (Clo) can be calculated from the following equations:

Clo = KeVd

Clo = KeVd

where Kc represents the first-order elimination rate constant out of aqueous humor, Vd is the apparent volume of distribution for the eye, Ko is the constant rate input into the anterior chamber, T is the time for the constant rate input, DIC is an intracameral dose, and AUCINF is the area (to infinity) under the aqueous humor concentration-time curve.

Each equation above depends on assumption that require experimentation to be carefully planned. In Eq. (4), Ke can be easily obtained. It is calculated from the latter linear slope of the logarithm of drug concentration in aqueous humor measured over time or obtained from any other tissue concentration in distribution equililbrium with aqueous humor. The Vd term in Eq. (3) has been correctly determined for the eye by either of two methods. One method, the topical infusion technique (28,37,52,76), has been discussed previously and is the basis for Eq. (5), also shown in Figure 4. The other method for determining Vd is based upon measuring drug concentration in aqueous humor over time following an intracameral injection of a very small volume (5 mL) of drug solution. Whenever an intracameral injection is made, there is concern that drug elimination can be altered because of a breakdown of the blood-aqueous barrier. However, Mayers, Miller, and coworkers (4,5,88,89), as well as Tang-Liu and coworkers (2,3), have established that for intracameral injections of 5 mL solutions containing various drugs, no significant breakdown of the blood-aqueous barrier had occurred since protein concentration was < 1 mg/mL. Equation (6), sometimes referred to as a "dose-area" determination, was used by Tang-Liu et al. (2,3) to calculate Vd for flurbiprofen and levobunolol.

Table 6 contains Clo values for those drugs for which accurate determinations have been made. Values range from 13 to 28.7 mL/min, which are 2.8-6.1 times higher than aqueous humor clearance, suggesting pathways of elimination other than aqueous turnover. The two most likely alternate routes of elimination are metabolism and systemic uptake by the vascular tissues of the anterior uvea. However, accumulation and retention by the lens (over the time course of the experiment), as well as back diffusion into the cornea and tears followed by subsequent drainage, are all minor routes that are not likely to significantly contribute to ocular clearance values.

0 0

Post a comment