Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Investigative Ophthalmology & Visual Science, Vol. 31, No. 5, May 1990 Copyright © Association for Research in Vision and Ophthalmology Regional Ocular Gentamicin Levels after Transcorneal and Transscleral Iontophoresis Robyn E. Grossman, Douglas F. Chu, and David A. Lee Transcorneal and transscleral iontophoresis were compared to subconjunctival injection (control) in the delivery of gentamicin into rabbit eyes. Gentamicin levels in the cornea, aqueous, and vitreous were measured by a fluorescence polarization assay at various time intervals after treatment. A mean peak corneal concentration of 376.1 Mg/ml was achieved 2 hr after transcorneal iontophoresis. This was significantly higher than the level obtained in control eyes (P = 0.016). A mean peak aqueous humor concentration of 54.8 Mg/m' occurred 2 hr after transcorneal iontophoresis. This was significantly higher than the peak level of 14.2 Mg/ml after subconjunctival injection (P = 0.003). Inhibitory levels (approximately 5 Mg/ml) were maintained in both aqueous and cornea for 8 hr after transcorneal iontophoresis. After transscleral iontophoresis, the mean peak vitreous humor concentration was 53.4 Mg/ml at 16 hr and remained inhibitory through 24 hr; the peak aqueous level was 23.2 Mg/ml and remained inhibitory for 24 hr. Peak drug concentrations in the vitreous were significantly higher than control (P = 0.026). Therapeutic vitreous humor levels were not achievable after transcorneal iontophoresis or subconjunctival injection. Potential corneal toxicity of transcorneal iontophoresis was demonstrated by measuring corneal thickness and endothelial cell counts prior to and 3 days after transcorneal iontophoresis of gentamicin and balanced saline solution (BSS) (control). No significant differences existed between eyes treated with gentamicin compared to those treated with BSS or when pre- versus postiontophoresis of gentamicin in the same eyes were compared. Transcorneal and transscleral iontophoresis may be an effective noninvasive method of delivering inhibitory levels of gentamicin to the cornea, aqueous humor, and vitreous for the treatment of intraocular infections. Invest Ophthalmol Vis Sci 31:909-916,1990 Iontophoresis is a technique of introducing drugs in the form of ions into tissues noninvasively, by the means of an electric current.1 This technique has been used in several areas of medicine—for example, in the administration of local anesthesia for myringectomy,2 as a means of inducing localized sweating with pilocarpine for diagnostic sweat testing in patients suspected of having cystic fibrosis,3 and in the administration of vidarabine monophosphate to patients with herpes orolabialis.4 In dentistry, fluoride has been iontophoresed in patients with hypersensitive dentin.5 Iontophoresis has several potential applications in ophthalmology, namely in achieving inhibitory levels of drugs in the eye for the treatment of bacterial endophthalmitis and keratitis. To date, var- ious researchers have used a myriad of methods to achieve inhibitory levels of drugs in the cornea, aqueous, and vitreous humor of the eye. Subconjunctival, retrobulbar, intravenous, and intramuscular injections along with topical application all have been tried,6"10 but most do not achieve adequate drug levels and involve other complications as well. Iontophoresis has been studied as a noninvasive method of solving some of these difficulties in achieving inhibitory levels of drugs in ocular tissues. Among the drugs that deserve attention is the aminoglycoside gentamicin, which has two positive charges per molecule at physiologic pH and a molecular weight of approximately 430; is lipid insoluble; and is effective against a wide variety of pathogenic gram-negative and gram-positive bacteria.""13 Previous gentamicin studies involving Pseudomonas have determined the mean inhibitory concentration to be 3.13 /u,g/ml14; another study showed that 94% of 1142 isolates were sensitive to 10 Mg/ml or less.9 In addition, gentamicin concentrations of 5 yug/ml or less inhibit over 90% of strains of Proteus rettgeri, P. vulgaris, and P. marganii,i5 whereas 99% of Staphylococcus strains are sensitive to 5 /xg/ml or less.16 Gentamicin is toxic at levels above 285 /tg/ml.17 Because of gentamicin's From the Jules Stein Eye Institute, Department of Ophthalmology, University of California—Los Angeles School of Medicine, Los Angeles, California. Supported by Karl Kirchgessner Foundation, Lucille Ellis Simon Research Fund, Elsie B. Ballantyne Research Fund, Research to Prevent Blindness, and National Eye Institute grant EY-07701. Submitted for publication: January 5, 1989; accepted September 13, 1989. Reprint requests: David A. Lee, MD, Jules Stein Eye Institute, Room 2-118, 800 Westwood Plaza, Los Angeles, CA 90024-1771. 909 Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 910 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / May 1990 polar nature, small molecular weight, and lipid insolubility, it is well suited for use in iontophoresis. Transcorneal and transscleral iontophoresis each have potential uses in the treatment of disorders involving different areas of the eye. Investigators have been able rapidly to achieve high levels of drug in the cornea and aqueous humor after transcorneal iontophoresis.1418"20 Others have been able to achieve high levels of drug in the aqueous and vitreous humor after transscleral iontophoresis.21"26 The purpose of the current study is to compare transcorneal and transscleral iontophoresis to each other and to subconjunctival injection in achieving regional ocular gentamicin concentrations at various time intervals after treatment. Materials and Methods Gentamicin sulfate powder (Sigma, St. Louis, MO) was combined with 2% agar to form a concentration of 100 mg/ml as follows: 0.2 g agar (Sigma) was dissolved in 7.0 ml of distilled water heated to near-boiling. The mixture was allowed to cool to approximately 50°C, and 1.0 g gentamicin sulfate powder was added. Enough distilled water was then added to this mixture to make a total volume of 10.0 ml. The caps were removed from 0.25 ml microcentrifuge tubes (Fisher Scientific, Pittsburgh, PA), and these tubes were filled with the mixture from the bottom to within 0.25 inch from the top, with care to avoid air bubbles. The tubes were then recapped, dated, and refrigerated at 20°C. When ready to use, the distal end of the microcentrifuge tube was cut off with a razor blade. A fresh tube was used for each eye. The internal diameter of the end of the tube was approximately 3 mm. The agar was then extruded approximately 2 mm by inserting the tube firmly onto a tuberculin syringe. The anode of the iontophoresis device was then inserted through the syringe into the agar. The device was grounded by attaching the other electrode by means of an alligator clip to the shaven ear contralateral to the eye receiving iontophoresis. Healthy adult pigmented rabbits weighing 2-3 kg were used in this study. The study conformed to the ARVO Resolution on the Use of Animals in Research. Immediately before iontophoresis the animals were anesthetized with intramuscular ketamine HC1 60 mg/kg (Parke Davis, Morris Plains, NJ) and intramuscular Xylazine 12 mg/kg (Mobay, Shawnee, KS). The eyes were then proptosed with a lid speculum. For transcorneal iontophoresis, the agar-gentamicin electrode was applied to the central cornea of one eye with a current of 0.2 mA for 10 min. The iontophoresis device used was based on a design by Brubaker.27 At the end of this period of time, the eye was Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 Vol. 31 irrigated with balanced salt solution (BSS; Alcon, Fort Worth, TX) to remove any residual drug from the ocular surface. The fellow eye received a 20-mg subconjunctival injection of gentamicin sulfate (0.5 ml of 40 mg/ml; Elkin-Sinn, Cherry Hill, NJ). At 0.5, 1, 2, 4, 8, and 16 hr after transcorneal iontophoresis and subconjunctival injection, aqueous humor, vitreous humor, and corneal samples were taken in respective order. The aqueous humor was removed by paracentesis using a 25-gauge needle attached to a 3-cc syringe. The needle was inserted through the peripheral clear cornea into the anterior chamber, with care to avoid damaging the iris, lens, or cornea. Approximately 150 ix\ aqueous humor was aspirated. The vitreous humor was removed by inserting an 18-gauge needle through the pars plana 2-3 mm posterior to the limbus into the center of the vitreous cavity, and 0.5-1.0 ml vitreous was aspirated. To further liquefy the vitreous humor, it was expressed through a 25-gauge needle into a microcentrifuge tube. The corneal samples were obtained by trephining the cornea with an 8.5mm diameter trephine and then excising the central cornea with scissors. Once removed, the cornea button was rinsed with 5 drops of BSS (0.33 ml) and frozen on dry ice until prepared for assay. At that time the corneas were thawed and placed individually on a clean glass microscope slide. With a straightedged razor, the corneas were minced into very fine pieces and put into preweighed microcentrifuge tubes. The mincing and transfer process took approximately 30 sec per cornea. After reweighing the tubes to determine the weight of the mass of the cornea tissue, the tubes were agitated in a Vortex mixer for 10 sec. This ensured that no pieces of cornea were adherent to the sides of the tubes. To each tube, 0.5 ml 0.01 M phosphate buffered saline was added. The tubes were allowed to incubate for 18 hr in a water bath heated to 37 °C and shaking at 100 oscillations/ min. The tubes were then centrifuged for 10 min at 2000 rpm. The supernatant was pipetted and placed into separate microcentrifuge tubes. All samples were stored on dry ice before assaying. The concentration of gentamicin in the cornea was derived with the following formula: (cone gentamicin in cornea) _ (cornea wt + buffer vol) (cornea wt) X (cone gentamicin in buffer) Drug concentrations are expressed in micrograms per gram tissue or micrograms per milliliter buffer. Buffer volume is expressed in milliliters, and cornea weight in grams.28 IONTOPHORESIS OF GENTAMICIN / Grossman er al No. 5 911 rabbits both before and 3 days after iontophoresis. One eye of each rabbit had transcorneal iontophoresis of gentamicin as described above. The fellow eye was used as a control and received transcorneal iontophoresis of BSS under the same conditions. As measures of corneal toxicity, two variables were calculated: increase in corneal thickness and decrease in endothelial cell count. A comparison was made for the two variables between the eyes treated with BSS and those treated with gentamicin; in addition, values after iontophoresis were compared to the values before iontophoresis for both groups of eyes. Data analysis included gentamicin levels in aqueous, vitreous, and cornea in transcorneal iontophoresis compared to subconjunctival injection, and in transscleral iontophoresis compared to subconjunctival injection. A Wilcoxon rank sum test was used with the accepted level of significance at P < 0.05. For the toxicity studies, the analysis of variance (ANOVA) test was used to determine if there was a difference between the corneas treated with gentamicin and those treated with BSS, as well as to compare preiontophoresis versus postiontophoresis values. Transscleral iontophoresis was performed with a new device which was similar to the device used for transcorneal iontophoresis but which could deliver higher adjustable current levels. Transscleral iontophoresis with this device was performed in a manner similar to transcorneal iontophoresis, except that the current was adjusted to deliver 2.0 mA (28.2 mA/cm2) for 10 min, and the electrode was placed 2-3 mm posterior to the limbus over the pars plana area of the sclera. The fellow eye received a 20-mg subconjunctival injection of gentamicin sulfate as described above. At 0.5, 1, 2, 4, 8, 16, 24, and 36 hr after transscleral iontophoresis and subconjunctival injection, aqueous humor and vitreous humor samples were removed as described above. All aqueous humor, vitreous humor, and cornea samples were assayed for gentamicin levels by the University of California—Los Angeles Toxicology Laboratory using a fluorescence polarization assay (TDx System Analyzer; Abbott Laboratories Diagnostic Division, Irving, TX). In addition, the gentamicin-agar mixture was assayed in order to ensure that the drug had not been inactivated during the heating process which incorporated gentamicin into the liquid agar. The potentially toxic effects of transcorneal iontophoresis on the cornea were monitored by measuring corneal thickness and endothelial cell counts with a Heyer Schulte specular microscope (Model HS-CEM 3; Coopervision, Irvine, CA), with Kodak Tri-X black and white film (ASA-400). The endothelial cell count for each eye was the average of at least three separate counts in central corneal areas of 0.02 mm2. These measurements were made in both eyes of six Results Table 1 summarizes the mean gentamicin levels in the vitreous and aqueous after transscleral iontophoresis, and in the vitreous, aqueous and cornea after both transcorneal iontophoresis and subconjunctival injection. The peak gentamicin level in the cornea after transcorneal iontophoresis was 376.1 jug/ml, achieved 2.0 hr after treatment. This was significantly higher Table 1. Regional gentamicin concentrations after iontophoresis and subconjunctival injection Hr Region Treatment Aqueous subconj. humor injection transcorneal iontophoresis transscleral iontophoresis Cornea 0.5 1 2 4 8 16 24 8.2 ± 5.3 (10) 32.0 ± 5.4* (10) 20.8 ± 12.7* 12.4 ± 6.0 (15) 41.8 ± 10.3* (12) 18.7 ± 8.8 14.2 + 4.5 (5) 54.8 + 21.7* 2.0 ± 0.9 0.2 ± 0.1 — (10) 4.7 ± 1.4* (5) 8.9 ± 4.6* (6) 0.4 ± 0.2 — 23.2 ± 14.3 8.1 ± 4.5 (14) 21.9 ± 6.0* (8) 7.9 ± 4.3 (8) (8) (7) (6) (5) (4) 5.9 ± 3.3 (4) — — 0.0 ± 0.0 (4) — — — (9) 20.8 + 10.2 15.9 ± 7.2 subconj. 84.7 ±31.8 71.8 ± 63.5 injection (4) (4) (4) (3) transcorneal 323.8 ± 84.3* 283.0 ± 185.6* 376.1 ± 206.7* 48.1 ± 10.7 iontophoresis (6) (9) (4) (6) Vitreous subconj. humor injection transcorneal iontophoresis transscleral iontophoresis 2.6 ± 2.5 (6) 0.5 ± 0.5 (4) 6.8 ± 2.9* (6) 2.2 ± 3.8 (6) 0.4 ± 0.2 (4) 8.3 ± 8.6* 0.8 + 0.3 (4) 0.3 + 0.2 (3) 12.7 + (6) All drug concentrations given in ng/m\; mean ± standard deviation. Number in parentheses is n. Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 (6) 0.4 ± 0.6 (6) 0.4 ± 0.2 (4) (2) 7.5 ± 4.9* 8.0 ± 4.5 (3) 11.6 ± 2.7 (4) 0.4 ± 0.3 (6) 0.2 ± 0.1 (3) 9.2* 25.3 ± 17.5* 37.5 ± 18.7* 53.4 ±35.4* 46.4 ±12.6* (4) (3) (5) (3) * Iontophoresis drug level is significantly higher than control (subconjunctival) level, P < 0.05. 912 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / May 1990 than the mean level of 20.8 Mg/ml, achieved in control eyes at 2.0 hr after a subconjunctival injection of gentamicin (P = 0.016). Mean corneal concentrations after transcorneal iontophoresis were significantly higher (P < 0.05) at 0.5, 1, and 2 hr than those in control eyes (Fig. 1). Gentamicin levels in the cornea remained inhibitory through 8 hr after both transcorneal iontophoresis and subconjunctival injection. The peak gentamicin concentration in the aqueous humor was 54.8 ^g/ml 2 hr after transcorneal iontophoresis, which was almost four times higher than the peak level of 14.2 /ig/ml achieved in control eyes. Aqueous humor levels after transcorneal iontophoresis were significantly higher than levels in control eyes for all time intervals measured except 16 hr (Fig. 2). Significant differences between aqueous humor levels after transscleral iontophoresis and control eyes existed only 0.5, 8 and 16 hr after treatment (P = 0.018, P = 0.003, and P = 0.013, respectively). After transscleral iontophoresis, however, the aqueous humor levels were still within inhibitory range at 24 hr. Transscleral iontophoresis resulted in very high concentrations of gentamicin in the vitreous humor, Vol. 31 and inhibitory levels were achieved as early as 0.5 hr after treatment (Fig. 3). The levels continued to rise steadily, reaching a peak of 53.4 jig/ml at 16 hr, and remained inhibitory through 24 hr. At 36 hr no drug could be detected in the vitreous humor. Peak vitreous levels after transscleral iontophoresis were over 20 times higher than the peak level achieved by subconjunctival injection. Transscleral iontophoresis was shown to achieve significantly higher values (P < 0.05) than subconjunctival injection at all measured time intervals. (Gentamicin levels were assumed to be undetectable in the vitreous 24 hr after subconjunctival injection.) Inhibitory vitreous humor levels were never achieved after transcorneal iontophoresis or subconjunctival injection, and there was no significant difference between these two methods at any time interval. In assessing potential toxicity to the cornea, we found that after iontophoresis the mean increase in corneal thickness from pretreatment values (0.320.35 mm) was not significantly different from zero for either the control eyes (-0.007 mm) or the gentamicin eyes (0.012 mm). The mean decrease in corneal cell count after treatment was actually larger for the 400 -i a FOLLOWING TRANSCORNEAL IONTOPHORESIS •— FOLLOWING SUBCONJUNCTIVAL INJECTION 300 O _ Fig. 1. Gentamicin concentration in cornea after transcorneal iontophoresis and subconjunctival injection. 200 - 100- HOURS POST IONTOPHORESIS Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 IONTOPHORESIS OF GENTAMICIN / Grossman er ol No. 5 Q FOLLOWING TRANSCORNEAL IONTOPHORESIS •— FOLLOWING SUBCONJUNCTIVAL INJECTION ---»-- oc Fig. 2. Gentamicin concentration in aqueous after transcorneal iontophoresis, transscleral iontophoresis, and subconjunctival injection. LJJ LJJ (/) s FOLLOWING TRANSSCLERAL IONTOPHORESIS 2 H o z o u _ o 913 •H C a a> E • •* E o> 30 10 40 HOURS POST IONTOPHORESIS o— FOLLOWING TRANSSCLERAL IONTOPHORESIS •— FOLLOWING SUBCONJUNCTIVAL INJECTION ---m-— Fig. 3. Gentamicin concentration in vitreous after transcorneal iontophoresis, transscleral iontophoresis, and subconjunctival injection. HOURS POST IONTOPHORESIS Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 FOLLOWING TRANSCORNEAL IONTOPHORESIS 914 Vol. 31 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / May 1990 control eyes (317.0 cells/mm2) than for the gentamicin eyes (281 cells/mm2). The decrease in cell count was significantly different from zero for the control group but not for the gentamicin group. When analyzed to see if there was a difference between the eyes treated with gentamicin and those treated with BSS, we found that neither the mean increase in corneal thickness (P = 0.18) nor the decrease in corneal cell count (P = 0.81) were significantly different between the two groups. Is! Discussion <D Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 U <U E 6 E IP* 1=1 1 m 00 O O o o o o E- d 254. — VO 765. * 101. 254. 254. q o o (N O o ex o o 0.05 E E ecup, d = 10 These experiments show that transcorneal iontophoresis may be a safe and effective means of achieving inhibitory concentrations of gentamicin in the aqueous humor and cornea. These results are consistent with a previous study in which inhibitory levels of gentamicin were obtained in the cornea and aqueous humor after transcorneal iontophoresis.14 The gentamicin concentrations in the cornea and aqueous after transcorneal iontophoresis were also strikingly higher than those reported by Insler et al after topical and systemic administration.9 Corneal concentrations at approximately 1 hr after administration can easily be compared as follows: 283.0 /zg/ml after transcorneal iontophoresis, 71.8 /xg/ml after subconjunctival injection, 16.2 ^ig/g after topical administration (3 drops of 13.6 mg/ml), and 6.1 ^g/g after intramuscular injection (four injections of 1 mg/kg each).9 Aqueous humor concentrations, in respective order, are: 41.8 /ig/ml, 12.4 /xg/ml, 0.3 jug/ml, and 0.4 /ug/ml.9 Transcorneal iontophoresis is not, however, an effective means of achieving therapeutic drug concentrations in the vitreous humor in the phakic rabbits used in our study. It has been shown that inhibitory drug levels can be obtained in the vitreous humor after transcorneal iontophoresis in aphakic rabbits because of the absence of the lensiris barrier.14 Transscleral iontophoresis was attempted with our original device designed by Brubaker,27 but we were unable to achieve inhibitory concentrations of drug in the vitreous humor because the current (0.2 mA) was insufficient. A review of previous research using transscleral iontophoresis of gentamicin showed successful experiments using a current density of 63.7-765 mA/cm2 (Table 2). Table 2 summarizes the known literature of iontophoresis of gentamicin and compares the relationships between current level, current density, duration of iontophoresis, and iontophoresis location to regional drug concentrations. Our original device could deliver a maximum current density of only 2.82 mA/cm2. To solve this problem, we designed a new iontophoresis device in which O E b o c 73 c 2 2 E E No. 5 915 IONTOPHORESIS OF GENTAMICIN / Grossman er al current density could be increased to 28.2 mA/cm2. The new device proved quite effective in delivering sufficient drug concentrations for our study and was later confirmed effective in studies by Choi and Lee25 and Grossman and Lee.26 Our modified iontophoresis device differed from those used in previous studies21'24 in that we made use of a probefilledwith the drug solution rather than an eyecup. Previous authors made use of the eyecup held under negative pressure in order to prevent a buildup of bubbles, to prevent leakage of the drug solution, and to keep the iontophoresis solution firmly adhered to the eye. The agar-gentamicin mixture proved to be quite useful in preventing any of the solution from dripping out of the microcentrifuge tube during the iontophoresis, and also prevented bubble formation. In addition, we believed that firmly holding the probe against the eye would provide adequate contact. Further studies25'26 successfully made use of the modified iontophoresis box using a nonagar-based drug solution, and thus found it to be effective with liquids as well. Our technique may be easier to use and more convenient than previous iontophoresis techniques. From our studies on corneal toxicity of gentamicin when delivered by iontophoresis, we found no significant difference between changes in corneal thickness and endothelial cell counts in eyes treated with gentamicin and eyes treated with BSS. However, in four of six corneas treated with gentamicin, there were mild corneal opacities, which were not present in any of the six eyes treated with BSS. This suggests that the opacities were due to the gentamicin rather than to the electric current. Other investigators who have extensively studied potential corneal damage by electric current maintain that current densities of up to 20 mA/cm2 are safe in rabbits. The current used for transcorneal iontophoresis in our study was 2.82 mA/cm2 and was therefore well below the toxic level.18 We were not as concerned about damage caused to the retina by the current in eyes receiving transscleral iontophoresis because the electrode was applied over the pars plana, an area which is not critical for vision. In addition, it has been shown that the procedure is well tolerated using a much higher current density.23 In conclusion, iontophoresis is an effective and noninvasive means of delivering inhibitory concentrations of gentamicin to various portions of the eye. High corneal and aqueous humor levels are achieved rapidly after transcorneal iontophoresis, and this technique may have useful applications in the treatment of bacterial keratitis. Therapeutic and sustained vitreous humor levels are best achieved with transscleral iontophoresis. This technique shows promise as either a primary treatment or as an adjunct to Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 intravitreal injection for the treatment of bacterial endophthalmitis. Key words: gentamicin, iontophoresis, subconjunctival injection, transcorneal, transscleral Acknowledgments The authors wish to thank members of the University of California—Los Angeles (UCLA) Toxicology Laboratory for their help in performing assays. The authors also wish to thank Noel Wheeler and Susan Ingalls of the UCLA Department of Biostatistics for their expertise in performing statistical analysis. References 1. Erlanger G: Iontophoresis, a scientific and practical tool in ophthalmology. Ophthalmologica 128:232, 1954. 2. Bridger MWM, Keene M, Graham JM, Healy R, and Ammor MM: A device for iontophoretic anesthesia of the tympanic membrane. J Med Eng Technol 6:62, 1982. 3. Gibson LE and Cooke RE: A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics 23:545, 1959. 4. Gangarosa LP, Hill JM, Thompson BL, Leggett C, and Rissing JP: Iontophoresis of vidarabine monophosphate for herpes orolabialis. J Infect Dis 154:930, 1986. 5. Gangarosa LP and Park NH: Practical considerations in iontophoresis offluoridefor desensitizing dentin. J Prosthet Dent 19:173, 1978. 6. Golden B and Coppel SP: Ocular tissue absorption of gentamicin. Arch Ophthalmol 84:792, 1970. 7. Barza M, Kane A, and Baum JL: Regional differences in ocular concentrations of gentamicin after subconjunctival and retrobulbar injection in the rabbit. Am J Ophthalmol 83:407, 1977. 8. Barza M, Kane A, and Baum JL: Intraocular penetration of gentamicin after subconjunctival and retrobulbar injection. Am J Ophthalmol 85:541, 1978. 9. Insler MS, Helm CJ, and George WJ: Topical vs systemic gentamicin penetration into the human cornea and aqueous humor. Arch Ophthalmol 105:922, 1987. 10. Rubenstein E, Goldfarb J, Keren G, Blumenthal M, and Treister G: The penetration of gentamicin into the vitreous humor in man. Invest Ophthalmol Vis Sci 24:637, 1983. 11. Physician's Desk Reference, 42nd ed. Dradell, NJ, Medical Economics, 1988, p. 1910. 12. Windholz M and Budavari S, editors: The Merk Index, 10th ed. Rathway, NJ, Merck & Co, 1983, p. 4254. 13. Havener WH: Ocular Pharmacology, 4th ed. St. Louis, CV Mosby, 1978, p. 198. 14. Fishman PH, Jay WM, Rissing JP, Hill JM, and Shockley RK: Iotophoresis of gentamicin into aphakic rabbit eyes: Sustained vitreal levels. Invest Ophthalmol Vis Sci 25:343, 1984. 15. Kirby WM and Standiford HC: Gentimicin: In vitro studies. J Infect Dis 119:361, 1969. 16. Waitz JA and Weinstein MJ: Recent microbiological studies with gentamicin. J Infect Dis 119:355, 1969. 17. Zachary I and Forester R: Experimental intravitreal gentamicin. Am J Ophthalmol 82:604, 1976. 18. Hughes L and Maurice DM: A fresh look at iontophoresis. Arch Ophthalmol 102:1825, 1984. 19. Von Sallman L: Penetration of penecillin into to the eye. Arch Ophthalmol 34:195, 1945. 916 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Moy 1990 20. Rootman DS, Hobden JA, Jantzen JA, Gonzalez JR, O'Callaghan RJ, and Hill JM: Iotophoresis of tobramicin for the treatment of experimental Pseudomonas keratitis in the rabbit. Arch Ophthalmol 106:262, 1988. 21. Barza M, Peckman C, and Baum J: Transscleral iontophoresis of cefazolin, ticarcillin, and gentamicin in the rabbit. Ophthalmology 93:133, 1986. 22. Barza M, Peckman C, and Baum J: Transscleral iontophoresis as an adjunctive treatment for experimental endophthalmitis. Arch Ophthalmol 105:1418, 1987. 23. Barza M, Peckman C, and Baum J: Transscleral iontophoresis of gentamicin in monkeys. Invest Ophthalmol Vis Sci 28:1033, 1987. Downloaded From: http://iovs.arvojournals.org/ on 05/12/2017 Vol. 31 24. Maurice DM: Iontophoresis of fluorescein into the posterior segment of the rabbit eye. Ophthalmology 93:128, 1986. 25. Choi TB and Lee DA: Transscleral and transcorneal iontophoresis of vancomycin in rabbit eyes. J Ocul Pharmacol 4:153, 1988. 26. Grossman R and Lee DA: Transscleral and transcorneal iontophoresis of ketoconazole in the rabbit eye. Ophthalmology 96:724, 1989. 27. Brubaker RF: The flow of aqueous humor in the human eye. Trans Am Ophthalmol Soc 80:391, 1982. 28. Mondino BJ: Alternate and classical pathway components of complement in the normal human cornea. Arch Ophthalmol 98:346, 1980.