3 Fluorophotometry for Pharmacokinetic Assessment |
63 |
done in the rabbit model after in vitro tissue culture testing for viability. The eye bath technique provides the most accurate and repeatable estimate of corneal epithelial permeability and is the method of choice for safety testing in the rabbit model.
The next issue in experimental design of the eye bath technique is to define the protocol for applying the test solution. The protocol objective is to provide statistical data separation of the baseline, negative control, and test solution effect on corneal epithelium permeability. A dry eye tear-lubricating solution is intended to be applied to the ocular surface 7–8 times/day at a 2 h interval. Safety of the solution is best evaluated with an exaggerated application protocol to increase the exposure to the solution. The solution may be tested in normal eyes for safety by:
1. Eye bath application: The ocular surface is continuously covered with the test solution for 3–5 min.
2. Multiple drops for 5–15 min: Two drops applied to the ocular surface each 1–2 min for 5–15 min. This provides a relatively continuous exposure without an eye bath.
3. Multiple drops for 1–5 days: Two drops applied to the ocular surface each 30 min for 6 h/day. This provides an exaggerated clinical application within the confines of an 8 h work day.
4. Clinical application: One drop applied by the subject to the ocular surface 7–8 times/day at a 2 h interval for between 1 day and multiple weeks.
McCarey and Reaves (1997) investigated the effects on epithelial permeability of preserved tear-lubricating solutions and preservative-free solutions on the rabbit eye. They applied the artificial tear solutions as a 5-min bath and multiple drops for 1 and 5 days. Their protocol will be used to assess the safety of the artificial tear-lubricating solutions within the rabbit model. I would suggest the protocol for test solution application in the initial rabbit model safety testing should be with the eye bath technique. The technique is performed as in the following description with the Fluorotron Master (OcuMetrics). The toxicity of a test solution can be evaluated by:
(a)3-min Dose Test: Apply the test solution as one drop/30 s for 3 min, i.e., six drops. Manually blink the eyelids. Wait 1 min and measure epithelial permeability.
(b)3-min Bath Test: Bathe the cornea in the test solution in vivo by the following technique. Place the unanesthetized rabbit on paper towels. Stand behind the rabbit while holding its lids open in a cup-like position. Fill the cul-de-sac “cup” with room temperature test solution for 3 min. Add test solution as needed to keep the cornea covered. Make sure the eyelids are pulled away from the cornea to permit good bathing of the corneal epithelium. After 3-min wick off excess artificial tear solution with a tissue wipe and measure epithelial permeability.
3.4 Clinical Applications of Fluorophotometry
Selecting a technique for testing safety of an artificial tear solution in the human subject is more restricting because of the consideration for patient compliance and comfort. The following discussion should be understood before making a technique selection.
64 |
B.E. McCarey |
Dry eyes are often responsible for severe complaints of discomfort. The tear film is unstable and cannot provide permanent wetting of the external ocular surface, i.e., the corneal epithelium and conjunctiva. The use of artificial tear solutions may have prolonged adverse effects on the epithelium and conjunctiva. The main morphological structure of the epithelial diffusion barrier is the tight junctions on the epithelial layers. If the diffusion barrier suffers even minor damage its permeability to hydrophobic substances increases considerably even if no epithelial lesions are visible with the slit lamp. Gobbels and Spitznas (1991) compared the corneal epithelial permeability with fluorophotometry before treating dry eye patient with artificial tear solution and after 8 weeks of treatment of applications at least five times/day. The corneal epithelial permeability of dry eye patients has been shown to be 2.8 times greater than that in individuals without ocular disease. Local preservatives are known to cause toxic side effects on the corneal epithelium such as disruption of cell membranes and increase permeability (Ramselaar et al. 1988). The preservative-free treatment had a −37% Ke change. The preservative treatment had a +21% Ke change. Corneal epithelial permeability of patients using artificial tear solutions with benzalkonium chloride were greater than control eyes by 3.1 times and solutions preserved with chlorobutanol were increased 1.7 times (Gobbles and Spitznas 1989). The authors concluded that the benzalkonium chloride preservative further stressed the corneal epithelium in the dry eye patients. Chlorobutanol-preserved artificial tear solutions improved the epithelium as expressed by a decrease in the epithelial permeability.
Gobbels and Spitznas (1991) expanded their initial study to detect possible changes in the permeability of the corneal epithelium in dry eye patients treated with artificial tear solutions. The patients were asked to apply the prescribed artificial tear solution every 2 h for at least 6 h/day. Prior to treatment and after 8 weeks of treatment, fluorophotometry was used to measure corneal epithelial permeability. The effect of aqueous tear substitutes on the tear film stability generally does not exceed 60–120 min, even though the bulk of the instilled aqueous solution does not presses longer than 15–20 min (Bron 1985). Eight weeks after the beginning of treatment, the corneal epithelial permeability of patients treated with a tear solution of 1.4% polyvinyl alcohol with 0.5% chlorobutanol or a solution of 2% polyvinyl alcohol without preservative was reduced significantly (−44.9% and 43.4% respectively). However, patients treated with 2% polyvinyl alcohol with 0.005% benzalkonium chloride showed no significant change in corneal epithelial permeability after treatment. These observations were further supported in another group of dry eye patients after 6 weeks of treatment (Gobbels and Spitznas 1992).
Benzalkonium chloride affects the semi-permeable corneal epithelial layer in two ways. Benzalkonium chloride leads to disruption of the zonula occludens, which seal off the superficial epithelial cells. The benzalkonium chloride molecules are incorporated into the cellular membranes of the epithelial cells by their lipophilic chains, thus providing gates for ionic, aqueous substances to penetrate through the lipophilic membranes into the intracellular spaces (Cadwallader and Ansel 1965; van Zutphen et al. 1971; Pfister and Burstein 1976; Burstein 1984).
3 Fluorophotometry for Pharmacokinetic Assessment |
65 |
The benzalkonium chloride molecules are bound onto the corneal surface immediately after instillation, such that the preservative escapes rapid washout by the tear film (Burstein 1980). Even 9 days after installation, a single drop of 0.01% benzalkonium chloride residues can be detected in the rabbit corneal epithelium by a radiocarbon technique (Pfister and Burstein 1976). The half-life of benzalkonium chloride in the rabbit corneal epithelium is about 20 h (Champeau and Edelhauser 1986). Several applications daily will cause accumulations of the preservative in the tissue. The accumulation can lead to further destabilization of the compromised dry eye ocular surface. After instillation of multiple drops of 0.01% benzalkonium chloride solution (four drops/day) into rabbit eyes, the amount of benzalkonium chloride present in the corneal and conjunctival tissues increased while the overall percentages of breakdown products were reduced (Champeau and Edelhauser 1986). In concentrations of 0.001–0.1% benzalkonium chloride exposure leads to loss of epithelial cell membrane microvilli, disruption of intracellular connections, and finally to complete desquamation of the superficial cell layers (Burstein and Klyce 1977).
Prior to testing the artificial tear-lubricating solution in human subjects, a protocol must be defined with rabbits to demonstrate with an exaggerated multiple drop frequency a corneal epithelial permeability difference between artificial tear-lubricating solution while retaining subject comfort and safety. The variable of corneal disease, i.e., dry eye pathology, should be avoided as a complicating variable. A negative control can be used to set upper limits of acceptable disruption of the corneal epithelial permeability. Visine™ is a commercially available tear-lubricating solution for dry eye relief. The solution contains 0.01% benzalkonium chloride. The product label states “instill 1 to 2 drops as needed.” The rabbit model testing should parallel the human subject testing. There have been many reports in the literature that may be used to provide guidance in establishing the protocol.
Schalnus and Ohrloff (1990) applied 20 mL of 2% sodium fluorescein into the conjunctival sac in rabbits and humans. The corneal fluorescence was measured at 55 min after application in the rabbit and 45 min after application in the human. The rabbit cornea uptake of fluorescein was 7.6 times greater than that in the human cornea. The authors felt that permeability kinetics of test solutions in the rabbit model must be transferred to the human with caution. The rabbit corneal epithelial permeability to sodium fluorescein was measured by Araie and Maurice (1987) in vivo to be 30 times greater than in human corneal epithelium. Hughes and Maurice (1984) in vivo measurement of sodium fluorescein permeability across the rabbit cornea was 10 times the human epithelial permeability values in the literature. The explanation for the greater permeability in the rabbit cornea than the human cornea can be justified from the physiological difference in the epithelial cell gap junctions, measurement technique or unidentified issues, such as prevalence of preexisting epithelial defects in the rabbit cornea.
Schalnus and Ohrloff investigated the effects of preservatives on the rabbit epithelium by applying 20 mL drops seven times at intervals of 5 min. The fluorescein uptake was 4.9 times greater in eyes treated with 0.01% benzalkonium chloride than that in untreated normal rabbit eyes. Repeating the experiment with an application
66 |
B.E. McCarey |
rate of three 20 mL drops/day for 10 days resulted in no difference between normal untreated control eyes and treated eyes.
Ramselaar et al. (1988) applied a 0.01% benzalkonium chloride solution at the rate of a 50 mL drop five times at intervals of 2 min. The fellow eye received comparable dosage with a control solution. The application protocol caused sufficient alteration in corneal epithelial permeability to distinguish between the control and test eyes, p < 0.03. A test solution containing 0.01% benzalkonium chloride and 1.0% tetracaine hydrochloride caused an even greater increase in epithelial permeability as compared to control, p < 0.005. Reducing the five-drop protocol to twoor three-drop protocol resulted in no increases in permeability, p > 0.05 and p = 0.05. A cut off in effect relative to dose frequency was well demonstrated with a fourdrop protocol, p > 0.025.
Paugh et al. (1998) performed an exaggerated exposure protocol with preserved (0.01% benzalkonium chloride and 0.03% EDTA) and nonpreserved artificial tearlubricating solutions in human subjects with ocular pathology. The solution application protocol was to perform a 5-min application of one-drop six times at intervals of 1 min, or a Multiple-Day Application of one drop eight times/day for 3 or 7 days. In each protocol, the corneal epithelial permeability was determined by fluorophotometry (Paugh and Joshi 1992; Joshi et al. 1996). The 5-min application protocol (n = 8) demonstrated a slight mean increase (test/control = 1.45) in permeability for the eyes receiving the preserved tear solution. Neither the 3-day nor the 7-day application protocol caused a change in epithelial permeability when comparing the preserved and nonpreserved solutions to baseline data. This study provides information on the design of an exaggerated test application protocol. The acute application 0.01% benzalkonium chloride at the rate of six drops in 5 min did cause a measurable increase in epithelial permeability, but the effect was minimal. Clinical dose rates did not cause a measurable change in permeability.
3.4.1 Endothelial Cell Layer Permeability
and Aqueous Humor Turnover
A fluorophotometer technique can be used to assess corneal endothelial cell layer permeability and aqueous humor turnover rate while treating the rabbit/subject with a test substance. The technique only requires several topical drops of fluorescein to the ocular surface. The test substance may be applied for any daily schedule prior to the fluorophotometer technique. The following description outlines a rabbit experimental protocol.
Use New Zealand White rabbits (3–4 kg body weight), n = 6 per experimental group. Treat the eyes in accordance to the prescribed drug regimen of a predetermined treatment schedule, such as QID, etc., for a predetermined number of days. The following experimental schedule is suggested:
1. At 8:15 a.m., apply 5 mL of 10% sodium fluorescein in BSS to the corneal as four applications with 15-min intervals ending at 9:00 a.m.
3 Fluorophotometry for Pharmacokinetic Assessment |
67 |
Fig. 3.12 Four hours after topical drops of sodium fluorescein to the rabbit eye, there is a steady-state exchange of fluorescein between the cornea and the aqueous humor. The endothelial cell layer permeability and aqueous humor turnover can be calculated from the rate of fluorescein exchange
2. Fifteen minutes after the last application (9:15 a.m.), rinse the remaining fluoresceinfrom the surface of the eye and fur with 40 mL of BSS.
3. Wait 15 min after the rinse (9:30 a.m.), then start the drug applications of one drop each hour throughout the rest of the day.
4. Start data collection 4 h (1:00 p.m.) after the last fluorescein application. Collect two scans each hour. Continue for 4 h (4:00 p.m.). The data should yield a linear line when plotted on semi-log scale (Fig. 3.12). The total experimental duration will be 7.75 h (finish at 4:00 p.m.).
The rabbit body weight, corneal thickness, corneal radius, corneal diameter, and anterior chamber depth are needed for the calculation of endothelial permeability and aqueous flow. Estimated values for the anterior chamber volume (Va = 200 mL) and corneal volume (Vc = 87 mL) can be used. OcuMetrics (Mountain View, CA) (OcuMetrics 1995) provides software program to perform the necessary data extraction from the fluorescent plots and calculations to determine endothelial permeability (kc. ca*q, mm/min), where kc. ca is endothelial permeability coefficient and q is corneal thickness in mm. The aqueous flow (ko*Va, mL/min) algorithms are also presented in the software program:
P = − |
dc |
× |
dCc |
, |
d(Cc −Ca ) |
|
|||
|
|
dt |
||
where P is endothelial cell layer permeability, dc is mean corneal stroma thickness, Cc is corneal fluorescein concentration, Ca is aqueous humor fluorescein concentration, and dt is change in time.