- Email: [email protected]
Collagen shield delivery of ofloxacinto the human eye
Collagen shield delivery of ofloxacin to the human eye Michael J. Taravella, MD, Jennifer Balentine, MD, David A. Young, PhD, Patricia Stepp, MD ABSTRACT Purpose: To determine the ocular penetration of ofloxacin into the anterior chamber of the human eye when delivered by a presoaked collagen shield. Sening: University of Colorado School of Medicine, Denver, Colorado. Methods: This prospective randomized clinical study comprised 31 patients having cataract surgery. Patients were divided into 2 groups: the first received 3 preoperative drops of commercially available topical ofloxacin 0.3% given 10 minutes apart; the second had a collagen shield soaked in the same medication applied to the eye before surgery. Aqueous humor was extracted immediately before surgery for analysis. Reeults: Mean aqueous concentration was 287 ng/mL _ 69 (SEM) (range 40 to 1141 ng/mL) in the drops group and 957 +-. 189 ng/mL (range 214 to 2437 ng/mL) in the shield group. The difference was statistically significant (P < .005). The minimum inhibitory concentration (MIC) for selected ocular pathogens is between 500 and 4000 ng/mL. Conclusions: A collagen shield presoaked in commercially available topical ofloxacin and applied before surgery appears safe. The MICs for many common ocular pathogens were reached or exceeded. Further study is recommended to determine whether this method of infection prophylaxis is an acceptable substitute for subconjunctival injections of antibiotics. J Cataract Refract Surg 1999; 25:562-565 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
everal studies have evaluated the delivery and penetration of antibiotics such as tobramycin, gentamicin, and vancomycin into the rabbit eye using presoaked collagen shields. 1-4 Other studies have shown the safety of this system in the perioperative setting?-" However, there is a paucity of data on the actual penetration of antibiotics into the human eye with this method. Although tobramycin in a commercially avail-
S
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
able concentration has been shown to penetrate poorly into the human eye when delivered by a presoaked collagen shield, other antibiotics have not been studied. 9 Our study sought to determine whether penetration into the human eye of ofloxacin, a commercially available fluoroquinolone antibiotic, can be enhanced using collagen shields as a delivery system.
Patients and Methods Acceptedfbr publication November 16, 1998.
From the Departmenu of Ophthalmology (Taravella, Balentine, Stepp) and PreventiveMedicine and Biometrics (Young), University of Colorado School of Medicine, Denver, Colorado, USA. Reprint requests to MichaelJ. Taravella, MD, 4200 East 9th Avenue, B-204, Denver, Colorado80262, USA. 562
The study protocol was reviewed and approved by the Human Subjects Committee of the Institutional Review Board at the University of Colorado Health Sciences Center, Denver. Patients were recruited from the population scheduled to have cataract surgery at
J CATARACTREFRACT SURG--VOL 25, APRIL 1999
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
University Hospital, Denver. Exclusion criteria included a known allergy to pork, pork collagen, or fluoroquinolone antibiotics. The procedure was explained to potential study patients, who were then enrolled at the time of surgery after providing informed consent. Surgical technique comprised clear corneal incision phacoemulsification performed using topical anesthesia. The corneal incisions were self-sealing, and none required sutures. Four drops of topical tetracaine 0.5% and 2 drops of topically applied bupivacaine 0.75% were placed in the superior and inferior conjunctival fornices before the standard surgical preparation. Anesthesia was supplemented by intracameral unpreserved lidocaine 1% (0.25 cc). The experimental protocol differed from standard cataract extraction in the following way: 31 patients were randomized into 2 groups: Group A received a commercially available ofloxacin ophthalmic solution 0.3% (Ocuflox®), 1 drop every 10 minutes for a total of 3 drops during the hour preceding cataract surgery. Group B received a Bausch & Lomb Bio-Cor II corneal collagen shield presoaked in ofloxacin for 10 minutes and then applied to the eye for approximately 60 minutes before surgery (mean time 68 minutes; range 27 to 109 minutes). All patients received a standard dilating regimen beginning about 1 hour before surgery and consisting of 3 drops each of cyclopentolate 1%, phenylephrine hydrochloride 2.5% (Neo-Synephrine®), tropicamide 1% (Mydriacyl®), and flurbiprofen 0.03% given as sets 15 minutes apart and concomitantly with the drug-soaked shield or ofloxacin drops. At the beginning of surgery, approximately 0.1 mL of aqueous humor was extracted using a tuberculin syringe with a 30 gauge cannula through a small anterior chamber paracentesis. The surgery proceeded as usual. Both groups received a subconjunctival injection of approximately 0.2 to 0.3 cc of cephapirin (Cefadyl ®) (200 mg/mL) at the end of surgery for infection prophylaxis. All patients had intact corneal epithelium preoperatively. The aqueous humor samples were stored in a freezer at -70°C until analysis. The samples were analyzed for ofloxacin concentrations using high-pressure liquid chromotography. Using t tests, ofloxacin concentrations in the aqueous were compared between groups and to a standard minimum inhibitory concentration (MIC) for several common ocular pathogens.
Regression analysis was used to model the time course of aqueous ofloxacin concentration as a function of duration of shield application. All analyses were done using the SAS statistical software package (SAS STAT, version 6.12, SAS Institute). During the study, no adverse reactions or surgical complications occurred.
Results There were 16 patients in the drops group and 15 in the shield group. There were no significant betweengroup differences in sex or mean age (67.9 years +_ 2.6 [SEM] and 65.4 -+ 2.3 years, respectively). Ofloxacin concentrations were not available for 2 patients (1 in each group) because the aqueous sample was insufficient. Mean aqueous ofloxacin concentration was 287 + 69 ng/mL (range 40 to 1141 ng/mL) in the drops group and 957 + 189 ng/mL (range 214 to 2437 ng/mL) in the shield group. The aqueous concentration of ofloxacin was significantly greater in the shield group (t = -3.3; df = 16.4 by Satterthwaite's approximation; P < .005). The results are summarized in Table 1.
Discussion Ofloxacin is a fluoroquinolone antibiotic that exhibits broad-spectrum activity against many grampositive and gram-negative organisms. *° It has been shown that ofloxacin drops given topically penetrate the cornea and aqueous well, achieving drug levels exceeding the MIC for many common ocular pathogens. 11-13The ability of ofloxacin to penetrate the eye well may be attributable, at least in part, to its excellent solubility in the tear film. The broad spectrum of antibacterial activity, together with its ability to penetrate, makes this antibiotic a good choice for infection prophylaxis for cataract surgery. Our study found better penetration into the eye with a collagen shield than with drops alone. Two other studies of drops have shown drug levels similar to those we achieved with the shield, 12a~ while 1 study found notably lower concentrations, t* The latter study used a drop regimen close to the one we used, while the 2 former studies used more prolonged regimens beginning 4 to 24 hours before surgery. The delivery regimen we used was chosen to reflect clinically realistic condi-
J C A T ~ C T ~ F ~ C T SURGIVOL 25, APRIL 1999
563
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
Table 1.
Aqueous concentration of ofloxacin: drops versus shield.
~ •
...
" . . ~ . ~ : . : ' : " AqueousCQn~p..ntratlon .... ,Ot'.eam) . " - . : ; : ( ~ L ) . "
'-Shield ,P8tlent '
. Age -Ofeam)
(nghnL)
(Minutes).
1
55
66
1
84
214
35
2
54
1141
2
62
517
97
3
54
195
3
60
780
60
4
66
462
4
61
904
85
5
71
137
5
65
536
70
6
64
78
6
58
2193
107
7
71
323
7
80
406
53
8
71
40
8
66
513
50
9
77
283
9
60
1330
55
10
60
288
10
60
788
60
11
87
254
11
71
292
27
12
80
145
12
54
706
60
13
66
334
13
63
2437
77
14
79
403
14
72
1775
109
15
64
157
.
tions not dependent on patient compliance. Differences in the number of drops given and dilution with other medications may also account for disparities observed in the aqueous concentrations between our study and others. Drug levels obtained with the shield exceeded the MIC for many common ocular pathogens, but not for Pseudomonas aeruginosa or Streptococcuspneumoniae (Table 2). However, our data showed a significantly linear trend toward higher drug concentrations with increased shield contact time (increase in concentration of 184 -+ 6.3 ng/mL for each additional 10 minutes of application; P < .02, regression analysis) (Figure 1). Sampling 1 to 2 hours later would likely yield higher drug concentrations in the aqueous than we found. Bouchard and coauthors '3 showed that the maximum concentration of ofloxacin when given topically should occur about 2 hours after the last drop is given. It is therefore reasonable to assume that the aqueous sampling in our study may have occurred before the time at which peak drug levels were achieved. Our study showed a high degree of variability in drug concentration in the aqueous in both the drop and shield groups. Factors that contributed to this variability indude dilution by other drops, no attempt to control patient blinking (and therefore tear turnover), 564
AqueousConcerldlmtlon ShieldTime
.
.
.
different shield application times, and variability in the time between shield removal, surgical preparation, and aqueous sampling. The significance of each of these factors was not determined. All patients had normal epithelium as determined by slitlamp examination; thus, epithelial status should not have been a factor in the difference in drug concentration between patients within a given group. However, the shield group received topical anesthetic drops (tetracaine) before appli-
Table 2. In vitro mean inhibitory concentrations (MICgo) of ofloxacin for selected organisms." Organism
Orion©In MICN (ng/mL)
Bacillus species Corynebacterium species Escherichia coil Moraxella
500 2000 125 125
Pseudornonas aeruginosa
4000
Serratia species
1000
Staphylococcus aureus
500
Staphylococcus epidermidis
500
Streptococcus pneumoniae
2000
MICgo -- minimum inhibitory concentration *Adapted from reference 10
j CATARACTREFRACTSURG--VOL25, APRIL 1999
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
2SO0
first day postoperatively in all patients, no formal comparison of punctate keratitis between the drop and shield group was made. Further studies addressing the issue of toxicity of collagen shield drug delivery are recommended before adopting its routine use.
i ]500 ]GO0
500
R~rellco~
20
SO
80
l]O
Minutes
Figure 1. (Taravella) Ofloxacin concentrations versus minutes of shield contact. cation of the shield, which would be expected to enhance penetration somewhat. Collagen shields appear to be well suited for drug delivery in the perioperative setting. Presoaked collagen shields release most of the drug within the first 30 minutes to 1 hour after application. Individual drug characteristics (e.g., solubility in the tear film, partition coefficient in the corneal epithelium, and molecular weight) determine how quickly peak levels of watersoluble antibiotics occur in the aqueous. In general, however, peak levels occur within 2 hours of application of the shield for most antibiotics studied to date. '4 This implies that peak levels occur soon after surgery, when a high level of antibiotics would be most needed to clear the anterior chamber of bacteria remaining in the eye. The advantage of using a collagen shield instead of drops for infection prophylaxis include a potential healing effect on the corneal epithelium, no concerns with patient compliance, and the ability to leave the eye patch undisturbed for the first 24 hours after cataract surgery, with the shield in place to deliver medication. Collagen shields have advantages over the use of subconjuctival antibiotic injections as well. These include ease of administration, no risk of perforation or subconjunctival hemorrhage, and patient comfort. The latter is especially important when cataract surgery is done using topical anesthesia. Potential adverse reactions with collagen shield drug delivery include epithelial and endothelial toxicity. No patient in our series developed significant or persistent corneal edema. However, preoperative and postoperative specular endothelial cell counts were not performed. Although fluorescein staining was done on the J
1. Unterman SR, Rootman DS, Hill JM, Parelman JJ, et al. Collagen shield drug delivery: therapeutic concentrations of tobramycin in the rabbit cornea and aqueous humor. J Cataract Refract Surg 1988; 14:500-504 2. Phinney RB, Schwartz SD, Lee DA, Mondino BJ. Collagen-shield delivery ofgentamicin and vancomycin. Arch Ophthalmol 1988; 106:1599-1604 3. Chen CC, Takruri H, Duzman E. Enhancement of the ocular bioavailability of topical tobramycin with use of a collagen shield. J Cataract Refract Surg 1993; 19:242245 4. O'Brien TP, Sawusch MR, Dick JD, et al. Use of collagen corneal shields versus soft contact lenses to enhance penetration of topical tobramycin. J Cataract Refract Surg 1988; 14:505-507 5. Aquavella JV, Ruffini JJ, LoCascio JA. Use of collagen shields as a surgical adjunct. J Cataract Refract Surg 1988; 14:492-495 6. Marmer RH. Therapeutic and protective properties of the corneal collagen shield. J Cataract Refract Surg 1988; 14:496-499 7. Poland DE, Kaufman HE. Clinical use of collagen shields. J Cataract Refract Surg 1988; 14:489-491 8. Haaskjold E, OhrstrOm A, Uusitalo RJ, et al. Use of collagen shields in cataract surgery. J Cataract Refract Surg 1994; 20:150-153 9. Taravella M, Stepp P, Young D. Collagen shield delivery of tobramycin to the human eye. CLAO J 1998; 24:166-168 10. Osato MS, Jensen HG, Trousdale MD, et al. The comparative in vitro activity of ofloxacin and selected ophthalmic antimicmbial agents against ocular bacterial isolates. Am J Ophthalmol 1989; 108:380-386; erratum 1991; 112:478-479 11. Donnenfeld ED, Schrier A, Perry HD, et al. Penetration of topically applied ciprofloxacin, norfloxacin, and ofloxacin into the aqueous humor. Ophthalmology 1994; 101:902-905 12. Price FW Jr, Whitson WE, Gonzales J, Jqhns S. Corneal tissue levels of topically applied ofloxacin. J Cataract Refract Surg 1997; 23:898-902 13. Bouchard CS, King KK, Holmes JM. The kinetics of anterior chamber ofloxacin penetration. Cornea 1996; 15:72-75 14. Drug delivery from collagen shields. Ocular Ther Manage 1991; 2(5):1-14
CATARACTREFRACTSURG--VOL 25, APRIL 1999
~65
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
everal studies have evaluated the delivery and penetration of antibiotics such as tobramycin, gentamicin, and vancomycin into the rabbit eye using presoaked collagen shields. 1-4 Other studies have shown the safety of this system in the perioperative setting?-" However, there is a paucity of data on the actual penetration of antibiotics into the human eye with this method. Although tobramycin in a commercially avail-
S
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
able concentration has been shown to penetrate poorly into the human eye when delivered by a presoaked collagen shield, other antibiotics have not been studied. 9 Our study sought to determine whether penetration into the human eye of ofloxacin, a commercially available fluoroquinolone antibiotic, can be enhanced using collagen shields as a delivery system.
Patients and Methods Acceptedfbr publication November 16, 1998.
From the Departmenu of Ophthalmology (Taravella, Balentine, Stepp) and PreventiveMedicine and Biometrics (Young), University of Colorado School of Medicine, Denver, Colorado, USA. Reprint requests to MichaelJ. Taravella, MD, 4200 East 9th Avenue, B-204, Denver, Colorado80262, USA. 562
The study protocol was reviewed and approved by the Human Subjects Committee of the Institutional Review Board at the University of Colorado Health Sciences Center, Denver. Patients were recruited from the population scheduled to have cataract surgery at
J CATARACTREFRACT SURG--VOL 25, APRIL 1999
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
University Hospital, Denver. Exclusion criteria included a known allergy to pork, pork collagen, or fluoroquinolone antibiotics. The procedure was explained to potential study patients, who were then enrolled at the time of surgery after providing informed consent. Surgical technique comprised clear corneal incision phacoemulsification performed using topical anesthesia. The corneal incisions were self-sealing, and none required sutures. Four drops of topical tetracaine 0.5% and 2 drops of topically applied bupivacaine 0.75% were placed in the superior and inferior conjunctival fornices before the standard surgical preparation. Anesthesia was supplemented by intracameral unpreserved lidocaine 1% (0.25 cc). The experimental protocol differed from standard cataract extraction in the following way: 31 patients were randomized into 2 groups: Group A received a commercially available ofloxacin ophthalmic solution 0.3% (Ocuflox®), 1 drop every 10 minutes for a total of 3 drops during the hour preceding cataract surgery. Group B received a Bausch & Lomb Bio-Cor II corneal collagen shield presoaked in ofloxacin for 10 minutes and then applied to the eye for approximately 60 minutes before surgery (mean time 68 minutes; range 27 to 109 minutes). All patients received a standard dilating regimen beginning about 1 hour before surgery and consisting of 3 drops each of cyclopentolate 1%, phenylephrine hydrochloride 2.5% (Neo-Synephrine®), tropicamide 1% (Mydriacyl®), and flurbiprofen 0.03% given as sets 15 minutes apart and concomitantly with the drug-soaked shield or ofloxacin drops. At the beginning of surgery, approximately 0.1 mL of aqueous humor was extracted using a tuberculin syringe with a 30 gauge cannula through a small anterior chamber paracentesis. The surgery proceeded as usual. Both groups received a subconjunctival injection of approximately 0.2 to 0.3 cc of cephapirin (Cefadyl ®) (200 mg/mL) at the end of surgery for infection prophylaxis. All patients had intact corneal epithelium preoperatively. The aqueous humor samples were stored in a freezer at -70°C until analysis. The samples were analyzed for ofloxacin concentrations using high-pressure liquid chromotography. Using t tests, ofloxacin concentrations in the aqueous were compared between groups and to a standard minimum inhibitory concentration (MIC) for several common ocular pathogens.
Regression analysis was used to model the time course of aqueous ofloxacin concentration as a function of duration of shield application. All analyses were done using the SAS statistical software package (SAS STAT, version 6.12, SAS Institute). During the study, no adverse reactions or surgical complications occurred.
Results There were 16 patients in the drops group and 15 in the shield group. There were no significant betweengroup differences in sex or mean age (67.9 years +_ 2.6 [SEM] and 65.4 -+ 2.3 years, respectively). Ofloxacin concentrations were not available for 2 patients (1 in each group) because the aqueous sample was insufficient. Mean aqueous ofloxacin concentration was 287 + 69 ng/mL (range 40 to 1141 ng/mL) in the drops group and 957 + 189 ng/mL (range 214 to 2437 ng/mL) in the shield group. The aqueous concentration of ofloxacin was significantly greater in the shield group (t = -3.3; df = 16.4 by Satterthwaite's approximation; P < .005). The results are summarized in Table 1.
Discussion Ofloxacin is a fluoroquinolone antibiotic that exhibits broad-spectrum activity against many grampositive and gram-negative organisms. *° It has been shown that ofloxacin drops given topically penetrate the cornea and aqueous well, achieving drug levels exceeding the MIC for many common ocular pathogens. 11-13The ability of ofloxacin to penetrate the eye well may be attributable, at least in part, to its excellent solubility in the tear film. The broad spectrum of antibacterial activity, together with its ability to penetrate, makes this antibiotic a good choice for infection prophylaxis for cataract surgery. Our study found better penetration into the eye with a collagen shield than with drops alone. Two other studies of drops have shown drug levels similar to those we achieved with the shield, 12a~ while 1 study found notably lower concentrations, t* The latter study used a drop regimen close to the one we used, while the 2 former studies used more prolonged regimens beginning 4 to 24 hours before surgery. The delivery regimen we used was chosen to reflect clinically realistic condi-
J C A T ~ C T ~ F ~ C T SURGIVOL 25, APRIL 1999
563
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
Table 1.
Aqueous concentration of ofloxacin: drops versus shield.
~ •
...
" . . ~ . ~ : . : ' : " AqueousCQn~p..ntratlon .... ,Ot'.eam) . " - . : ; : ( ~ L ) . "
'-Shield ,P8tlent '
. Age -Ofeam)
(nghnL)
(Minutes).
1
55
66
1
84
214
35
2
54
1141
2
62
517
97
3
54
195
3
60
780
60
4
66
462
4
61
904
85
5
71
137
5
65
536
70
6
64
78
6
58
2193
107
7
71
323
7
80
406
53
8
71
40
8
66
513
50
9
77
283
9
60
1330
55
10
60
288
10
60
788
60
11
87
254
11
71
292
27
12
80
145
12
54
706
60
13
66
334
13
63
2437
77
14
79
403
14
72
1775
109
15
64
157
.
tions not dependent on patient compliance. Differences in the number of drops given and dilution with other medications may also account for disparities observed in the aqueous concentrations between our study and others. Drug levels obtained with the shield exceeded the MIC for many common ocular pathogens, but not for Pseudomonas aeruginosa or Streptococcuspneumoniae (Table 2). However, our data showed a significantly linear trend toward higher drug concentrations with increased shield contact time (increase in concentration of 184 -+ 6.3 ng/mL for each additional 10 minutes of application; P < .02, regression analysis) (Figure 1). Sampling 1 to 2 hours later would likely yield higher drug concentrations in the aqueous than we found. Bouchard and coauthors '3 showed that the maximum concentration of ofloxacin when given topically should occur about 2 hours after the last drop is given. It is therefore reasonable to assume that the aqueous sampling in our study may have occurred before the time at which peak drug levels were achieved. Our study showed a high degree of variability in drug concentration in the aqueous in both the drop and shield groups. Factors that contributed to this variability indude dilution by other drops, no attempt to control patient blinking (and therefore tear turnover), 564
AqueousConcerldlmtlon ShieldTime
.
.
.
different shield application times, and variability in the time between shield removal, surgical preparation, and aqueous sampling. The significance of each of these factors was not determined. All patients had normal epithelium as determined by slitlamp examination; thus, epithelial status should not have been a factor in the difference in drug concentration between patients within a given group. However, the shield group received topical anesthetic drops (tetracaine) before appli-
Table 2. In vitro mean inhibitory concentrations (MICgo) of ofloxacin for selected organisms." Organism
Orion©In MICN (ng/mL)
Bacillus species Corynebacterium species Escherichia coil Moraxella
500 2000 125 125
Pseudornonas aeruginosa
4000
Serratia species
1000
Staphylococcus aureus
500
Staphylococcus epidermidis
500
Streptococcus pneumoniae
2000
MICgo -- minimum inhibitory concentration *Adapted from reference 10
j CATARACTREFRACTSURG--VOL25, APRIL 1999
COLLAGEN SHIELD DELIVERY OF OFLOXACIN
2SO0
first day postoperatively in all patients, no formal comparison of punctate keratitis between the drop and shield group was made. Further studies addressing the issue of toxicity of collagen shield drug delivery are recommended before adopting its routine use.
i ]500 ]GO0
500
R~rellco~
20
SO
80
l]O
Minutes
Figure 1. (Taravella) Ofloxacin concentrations versus minutes of shield contact. cation of the shield, which would be expected to enhance penetration somewhat. Collagen shields appear to be well suited for drug delivery in the perioperative setting. Presoaked collagen shields release most of the drug within the first 30 minutes to 1 hour after application. Individual drug characteristics (e.g., solubility in the tear film, partition coefficient in the corneal epithelium, and molecular weight) determine how quickly peak levels of watersoluble antibiotics occur in the aqueous. In general, however, peak levels occur within 2 hours of application of the shield for most antibiotics studied to date. '4 This implies that peak levels occur soon after surgery, when a high level of antibiotics would be most needed to clear the anterior chamber of bacteria remaining in the eye. The advantage of using a collagen shield instead of drops for infection prophylaxis include a potential healing effect on the corneal epithelium, no concerns with patient compliance, and the ability to leave the eye patch undisturbed for the first 24 hours after cataract surgery, with the shield in place to deliver medication. Collagen shields have advantages over the use of subconjuctival antibiotic injections as well. These include ease of administration, no risk of perforation or subconjunctival hemorrhage, and patient comfort. The latter is especially important when cataract surgery is done using topical anesthesia. Potential adverse reactions with collagen shield drug delivery include epithelial and endothelial toxicity. No patient in our series developed significant or persistent corneal edema. However, preoperative and postoperative specular endothelial cell counts were not performed. Although fluorescein staining was done on the J
1. Unterman SR, Rootman DS, Hill JM, Parelman JJ, et al. Collagen shield drug delivery: therapeutic concentrations of tobramycin in the rabbit cornea and aqueous humor. J Cataract Refract Surg 1988; 14:500-504 2. Phinney RB, Schwartz SD, Lee DA, Mondino BJ. Collagen-shield delivery ofgentamicin and vancomycin. Arch Ophthalmol 1988; 106:1599-1604 3. Chen CC, Takruri H, Duzman E. Enhancement of the ocular bioavailability of topical tobramycin with use of a collagen shield. J Cataract Refract Surg 1993; 19:242245 4. O'Brien TP, Sawusch MR, Dick JD, et al. Use of collagen corneal shields versus soft contact lenses to enhance penetration of topical tobramycin. J Cataract Refract Surg 1988; 14:505-507 5. Aquavella JV, Ruffini JJ, LoCascio JA. Use of collagen shields as a surgical adjunct. J Cataract Refract Surg 1988; 14:492-495 6. Marmer RH. Therapeutic and protective properties of the corneal collagen shield. J Cataract Refract Surg 1988; 14:496-499 7. Poland DE, Kaufman HE. Clinical use of collagen shields. J Cataract Refract Surg 1988; 14:489-491 8. Haaskjold E, OhrstrOm A, Uusitalo RJ, et al. Use of collagen shields in cataract surgery. J Cataract Refract Surg 1994; 20:150-153 9. Taravella M, Stepp P, Young D. Collagen shield delivery of tobramycin to the human eye. CLAO J 1998; 24:166-168 10. Osato MS, Jensen HG, Trousdale MD, et al. The comparative in vitro activity of ofloxacin and selected ophthalmic antimicmbial agents against ocular bacterial isolates. Am J Ophthalmol 1989; 108:380-386; erratum 1991; 112:478-479 11. Donnenfeld ED, Schrier A, Perry HD, et al. Penetration of topically applied ciprofloxacin, norfloxacin, and ofloxacin into the aqueous humor. Ophthalmology 1994; 101:902-905 12. Price FW Jr, Whitson WE, Gonzales J, Jqhns S. Corneal tissue levels of topically applied ofloxacin. J Cataract Refract Surg 1997; 23:898-902 13. Bouchard CS, King KK, Holmes JM. The kinetics of anterior chamber ofloxacin penetration. Cornea 1996; 15:72-75 14. Drug delivery from collagen shields. Ocular Ther Manage 1991; 2(5):1-14
CATARACTREFRACTSURG--VOL 25, APRIL 1999
~65