- Email: [email protected]
Efficacy of diclofenac sodium ophthalmic solution versus placebo in reducing inflammation following cataract extraction and posterior chamber lens implantation
Efficacy of diclofenac sodium ophthalmic solution versus placebo in reducing inflammation following cataract extraction and posterior chamber lens implantation Manus C. KrafT, M.D., Robert G. Martin, M.D., Albert C. Neumann, M.D., Arthur J. Weinstein, M.D.
ABSTRACT One hundred forty-eight patients were enrolled in a randomized, prospective, placebo-controlled clinical trial evaluating the efficacy of diclofenac sodium (Voltaren Ophthalmic®) in reducing ocular inflammation following extracapsular cataract extraction with posterior chamber intraocular lens implantation. Eligible patients were enrolled and randomized (2:1 diclofenac:placebo) if the sum of anterior chamber cells plus flare one day postoperatively (baseline) was at least four. None of the patients received concomitant steroidal anti-inflammatory treatment. The 99 patients receiving diclofenac sodium had significantly greater improvement from baseline in summed flare plus cell score than the 49 placebo patients at two to five days and seven to nine days after baseline. Similarly, diclofenac sodium patients had significantly less post-baseline conjunctival erythema and ciliary flush than placebo patients. Significantly more diclofenac sodium patients than placebo patients showed moderate to marked improvement from baseline in overall assessment of inflammatory response. Forty-nine percent of placebo patients but only 17% of diclofenac patients were considered therapeutic failures (P < .001 ). By five to seven days, 82% of diclofenac sodium patients and 59% of placebo patients had corrected visual acuities of 20/40 or better (P < .001 ). There were no clinically important differences in mean intraocular pressure at any visit.
Key Words: anterior chamber reaction, diclofenac sodium, nonsteroidal antiinflammatory drug, ocular inflammation, post-cataract inflammation
Diclofenac sodium (Voltaren Ophthalmic®) is a potent, rapidly acting nonsteroidal anti-inflammatory drug (NSAID). The oral formulation of diclofenac sodium received Federal Drug Administration (FDA) approval in 1988 for use in treating rheumatoid arthritis, osteoarthritis, and ankylosing spondylitis. Clinical ophthalmologic trials have demonstrated that topical and oral administration of diclofenac sodium effectively reduces a variety of signs and symptoms of ocular inflammation following cataract extraction and intraocular lens (IOL) implantation. A placebo-controlled trial of 100 intracapsular cata-
ract extraction cases by Ronen and coauthors 1 showed significantly decreased levels of anterior chamber flare and cells in the diclofenac treated group compared with a group that did not receive di.clofenac. Both groups received routine postoperative topical steroids. In another study of intracapsular cataract extraction cases, topical diclofenac treatment was compared with dexamethasone treatment; the study demonstrated equivalent efficacy for diclofenac sodium and the steroid. 2 Diclofenac has also been shown to have relatively greater efficacy in reducing ocular inflammation in iritis and iridocyclitis cases than dexamethasone. 3• Other ste-
From Northwestern University and Kraff Eye Institute, Chicago, Illinois (Krajj), Carolina Eye Associates, P.A., and Medical Care International Ophthalmic Institute, Southern Pines, North Carolina (Martin), Neumann Eye Institute, Deland, Florida (Neumann), and Eye Associates of New Mexico, Albuquerque, New Mexico (Weinstein). Reprint requests to Manus C. Kraff, M.D., Kraff Eye Institute, 5600 West Addison Street, Chicago, Illinois 60634. 138
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
ciliary flush (see Appendix for explanation of clinical grades), the overall assessment of inflammatory response (assessing change from baseline; see Appendix), and the therapeutic failure rate. A patient was considered a therapeutic failure if ( 1) he or she was withdrawn from the study because of an adverse experience, (2) if he or she was withdrawn because of an unsatisfactory therapeutic response to treatment, at the discretion of the surgeon, or (3) if the sum of anterior chamber cells and flare was the same or greater than the one-day postoperative (baseline) level. Ophthalmic safety was evaluated by changes from baseline in visual acuity and lOP and by observing change in the slitlamp examination and funduscopy. Instances of ocular or systemic reactions and the need for concomitant prescription medications other than tobramycin were also evaluated. A single clinical examiner performed all assessments at all visits at each site. Two of the examiners were ophthalmologists and two were optometrists. All four clinical examiners were trained together before initiation of the study. Quantitative variables were measured with the same instrument under the same examination conditions for all patients at a site. For postsurgical SUBJECTS AND METHODS inflammatory variables, the examiners from each site Four surgeons at four clinical centers participated in were trained by a lead observer who examined practice this study. Patients of either sex over the age of 35 were patients jointly with the clinical examiners and trained potentially eligible if they were scheduled to have routine them in consistent evaluation. All slitlamp examinacataract extraction by phacoemulsification or the tions were performed under standardized conditions. planned extracapsular method, followed by the implan- The criteria for performing and scoring the slitlamp tation of a posterior chamber IOL. Screening exclusion examination variables are described in detail in the criteria included intraocular pressure (lOP) greater than Appendix. Analysis of efficacy was based on the change in the 24 mm Hg, ocular medications other than hypotensives, of anterior chamber cells and flare from baseline in sum previous surgery in the eye being operated on, history of the treatment and placebo groups. In addition, ratings of ocular inflammation, recent steroid use, allergy to conjunctival erythema, ciliary flush, and overall assessNSAIDs, and blindness or absence of the fellow eye. ment of inflammatory response were compared. In each Potentially eligible patients were screened at one day instance, the groups were compared using a weightedpostoperatively (baseline examination). Patients with a of variance least-squares estimate in a linear analysis sum of anterior chamber cells and flare of four or greater model for mean response, confirmed by Cochran-Man(see Appendix for explanation of clinical grades) were randomized to receive 0.1% diclofenac sodium ophthal- tel-Haenszel tests controlling for potential effects of mic solution or placebo (solution vehicle only). Two study center. The proportions of therapeutic failures in active drug recipients were randomized for each placebo each group were compared with Cochran-MantelHaenszel tests controlling for study center. 6 recipient. If a patient was withdrawn following a study visit, his Patients were instructed to instill one drop of the or her ratings at that visit were carried through for all masked solution four times daily beginning 22 to 34 subsequent visits. All analyses were conducted on an hours postoperatively and continuing for 14 days. Paintent-to-treat basis (see Results). Statistical significance tients were examined at one day postoperatively (baseat P s .05 for all comparisons. All tests were was set line) and at three to five days, seven to nine days, and 14 two-tailed. The sample size was calculated to provide to 16 days postoperatively. One drop of tobramycin ~0.80 for therapeutic failure rate, in the absence power ophthalmic solution 0.3% was administered four times daily on the first day postoperatively and continued at of data to calculate power for efficacy variables. the discretion of the surgeon. Tobramycin was administered at least five minutes after the study solution. RESULTS The primary efficacy endpoint for this study was the A total of 148 patients were enrolled: 99 received change in the sum of anterior chamber cells and flare from baseline. Secondary efficacy endpoints included diclofenac sodium ophthalmic solution and 49 received the clinical assessment of conjunctival erythema and the placebo. Eleven patients (eight diclofenac, three plaraid-controlled studies have established that the preoperative and postoperative topical administration of 0.1% diclofenac sodium solution significantly speeds the re-establishment of the blood-aqueous barrier after surgery. 4 •5 While corticosteroids are not specifically approved by FDA for treating postoperative ocular inflammation, they are widely administered after cataract extraction. Because the efficacy of these steroids in reducing ocular inflammation had not been formally established, the FDA required that the efficacy of new anti-inflammatory drugs be established by placebo-controlled studies. This report presents the results of a multi-site, doublemasked, placebo-controlled clinical trial comparing diclofenac sodium ophthalmic solution with placebo (solution vehicle alone) to determine the anti-inflammatory efficacy, safety, and tolerability of diclofenac sodium following cataract surgery. Topical diclofenac sodium ophthalmic (Voltaren Ophthalmic®) solution received FDA approval for the treatment of post-cataract extraction inflammation in March 1991, based in part on this study.
J CATARACT !WFRACT SURG-VOL 20, MARCH 1994
139
cebo; 7% of the enrollment) were protocol violators with 33 visit violations (21 diclofenac; 12 placebo) of a total of592 planned postoperative visits. Violations included unallowed concomitant medication, excluded preoperative ocular pathology, occasional clinical assessments not performed by independent observers, elevated baseline lOP, and mildly out of range postoperative visits. Because of the low number of protocol violations, and no apparent group differences with respect to protocol violations, all enrolled cases were analyzed on an intentto-treat basis. The patients were predominantly white (95%) and female (61 %). The mean age was 70.6 years (standard deviation 9.3). A total of 26 variables were tested for baseline comparability. Only race differed significantly between the diclofenac (99% white, 1% other) and placebo (88% white, 12% other)groups(whitevs. other, P= .006 by Fisher's exact test), but race was not associated with any treatment response. Most patients had phacoemulsification (74%) rather than extracapsular extraction of their cataracts. At each site, the ratio of phacoemulsification to planned extracapsular extraction was similar in drug and placebo recipients. The method of cataract extraction was not associated with any statistically significant difference in any endpoint examined. Only one patient in each group experienced a posterior capsule rupture and none experienced vitreous loss or posterior capsulotomy. One patient in the drug group experienced a small inferior opening in the capsule and one suffered a partial iris prolapse. No patient had more than 50% hyphema after surgery. Four recipients of diclofenac sodium and 17 recipients of the placebo were prematurely withdrawn from the study. Reasons for withdrawal included adverse reactions (one diclofenac sodium recipient, two placebo recipients), unsatisfactory therapeutic response (three and 12 patients, respectively), and protocol or schedule violations (three placebo recipients). One patient experienced an adverse reaction at the first post-baseline examination prior to diclofenac treatment and was not included in efficacy analyses but was retained in safety analyses. Scores for anterior chamber cells and for anterior chamber flare may range from 0 to 4, and the sum of anterior chamber cells and flare may range from 0 to 8. All patients had a score of 4 or greater at baseline (one day postoperatively), which was the criterion for admission to the study. Tables 1 through 3 present the mean postoperative values for cells, flare, and the sum of cells and flare for patients receiving diclofenac sodium and placebo. Only one comparison, anterior chamber cells at three to five days postoperatively, failed to achieve statistical significance. Figure 1 is a line graph of the mean sum of cells and flare at baseline and at each postoperative time period. While the baseline starting scores were equal in both treatment groups, patients receiving diclofenac sodium 140
Table 1. Change from baseline in mean anterior chamber cell count. Time Period N Diclofenac Placebo
98 49t
3-5 Days 7-9 Days Mean (SE)* Mean (SE)
14-16 Days
1.44 (0.11) 0.80 (0.15)
2.16 (0.10) 1.06 (0.16)
0.70 (0.10) 0.48 (0.14)
Mean (SE)
Significance (P)
Source of variation
0.410 0.186
Drug Drug controlled for site
<0.001 0.001
<0.001 <0.001
* Standard error of the mean
t N = 48 at the 3-5 day visit
Table 2. Change from baseline in anterior chamber flare score. Time Period N Diclofenac Placebo
98 49t
3-5 Days
7-9 Days Mean (SE)* Mean (SE)
14-16 Days
0.49 (0.09) 0.17 (0.11)
1.48 (0.09) 0. 71 (0.14)
Drug controlled for site
Mean (SE)
Significance (P)
Source of variation Drug
0.91 (0.09) 0.35 (0.13)
<0.001 0.026
<0.001 0.001
<0.001 <0.001
* Standard error of the mean
t N = 48 at the 3-5 day visit
had lower mean scores than patients receiving the placebo solution at each time period after baseline. The distribution of summed flare and cell scores shifted from a median value of 7 at baseline to 5 at three to five days, to 4 at seven to nine days, and to 2 at 14 to 16 days in the diclofenac group. Patients receiving diclofenac sodium had greater improvement from baseline in summed flare and cell score than patients receiving placebo at all three post-baseline examinations (Table 3). Significant site differences were noted in the scores for anterior chamber cells, anterior chamber flare, and the sum of flare and cells. There was only one statistically significant interaction between drug and site, for anterior chamber flare at three to five days postoperatively (P :S .02). The differences between drug and placebo groups were statistically significant at all sites, evaluated individually, and the overall difference between drug and placebo groups remained statistically significant when adjusted for the differences among sites.
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
Table 3. Change from baseline in sum of cells and flare. Time Period 7-9 Days 3-5 Days Mean (SE)* Mean (SE) 1.19 (0.17)t 2.35 (0.18)
N
Diclofenac Placebo
98 48
1.14 (0.25)
0.65 (0.22)
14-16 Days Mean (SE) 3.64 (0.17) 1.78 (0.27)
Significance (P)
Source of variation
<0.001 <0.001 Drug 0.016 <0.001 <0.001 Drug 0.052 controlled for site * Standard error of the mean t Computed as (baseline sum of flare and cells) - (postoperative sum of flare and cells). Positive values represent improvements from baseline.
Figures 2 and 3 are graphs of mean conjunctival erythema and mean ciliary flush versus time for patients receiving diclofenac sodium or placebo. For both measures, the means were significantly lower for patients receiving diclofenac sodium than for those receiving the placebo at all post-baseline examinations. For all comparisons, the effect of diclofenac sodium, controlled for study site, was significant at P s .001 (Tables 4 and 5). Overall assessment of inflammatory response was graded as moderate to marked improvement over baseline or the same or worse than baseline. Figure 4 presents the proportions of cases with "moderate to marked improvement" and "same or worse than baseline" by treatment group for each postoperative period. Significantly more recipients of diclofenac sodium than recipients of placebo displayed moderate to marked improvement at seven to nine days and at 14 to 16 days relative to baseline (P ::5 .001 ). Significantly more recipients of placebo than of diclofenac sodium were rated as being
the same or worse than at baseline (P s .01 at all three periods). The rates of therapeutic failures are presented in Figure 5. At seven to nine days postoperatively, 49% of placebo recipients and 17% of diclofenac sodium recipients were considered therapeutic failures (P ::s; .001 ). At 14 to 16 days postoperatively, 39% of placebo recipients but only 4% of diclofenac sodium recipients were therapeutic failures (P ::5 .001). At two to five days postoperatively, 67% of diclofenac sodium recipients had best corrected visual acuity of 20/40 or better compared with 53% of placebo recipients (Figure 6). By five to seven days, the difference (82% diclofenac vs. 56% placebo) was statistically significant (P s .001), and it remained significant at 14 to 16 days postoperatively (84% vs. 59%, P s .001). At the final study (endpoint) visit, 80% of diclofenac sodium recipients had a two-line improvement in best corrected vision relative to their presurgery baseline, while only 57% of placebo recipients had a two-line improvement (P ::s; .0 1). The possible confounding effect of phacoemulsification versus planned extracapsular extraction on visual acuity was examined. Extraction technique was not associated with a statistically significant difference in visual acuity, nor was a statistically significant interaction with drug treatment found. There were no clinically important differences between the groups in the frequency or severity of postoperative changes in lOP at any visit. Mean endpoint lOP was 16.6 mm Hg for the diclofenac group and 15.8 mm Hg for the placebo group. Mean change from baseline lOP after two to five treatment days was -2.7 mm Hg for diclofenac and -2.8 mm Hg for placebo. Four glaucoma patients received diclofenac sodium and one received placebo. There were no adverse effects on glaucoma control. There were no significant differences between the groups in the frequency of adverse outcomes. One pa-
8 F 7 L A 6
R E
5
+
4
c
3
E L L
Fig. 1. **
2
s
0 0
2
3
4
5
6
8
7
9
10
11
Days On Treatment ----- Diclofenac
~ Placebo
12
13
14
15
(Kraft) Mean sum of cells and flare at baseline and each postoperative period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. * = P ~ .052; ** = P < .05.
16
I
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
141
c 0
4~-------------------
n j
Fig. 2.
(Kraft) Mean conjunctival erythema score at baseline and each postoperative period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. Erythema: 0 = none; 1 = pale red; 2 = red; 3 = bright red; 4 = definite chemosis. * = p < .01.
u n t i v
3
a
2
I
.----------~~
E
r y t h
e
oL-~~-L~--L-~~~~--L-~~~~--L-~-L~~~
m
0
a
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
I
Days On Treatment
I ___,._ Diclofenac
--+-- Placebo
4~----------------------------------------------,
c I L I
Fig. 3.
3
(Kraft) Mean ciliary flush score at ~ baseline and each postoperative y period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. Per- u centage of perilimbal region cov- S ered: 0 =none; 1 = 1%-25%; 2 = H 26%-50%; 3 = 51%-75%; 4 = 76%-100%. * = p < .01.
•
2
E
oL_~~_L~--L_~-L~~--L_J__L_J
0
1
2
3
4
5
6
7
8
_ _L_J__L~~L_~~
9 10 11 12 13 14 15 16 17 18 19 20
Days On Treatment ___,._ Diclofenac
Table 4. Mean difference* in conjunctival erythema.
Diclofenac Placebo Source of variation
N 98 48
Time Period 3-5 Days 7-9 Days Mean (SE)t Mean (SE) 0.66 (0.08) 1.38 (0.09) -0.08 (0.13) 0.02 (0.15)
14-16 Days Mean (SE) 1.73 (0.09) 0.39 (0.16)
Significance (P)
Drug <0.001 <0.001 <0.001 Drug <0.001 <0.001 <0.001 controlled for site *Computed as (baseline) - (postoperative). Positive values represent improvements from baseline. t Standard error of the mean
142
--+-- Placebo
I
Table 5. Mean difference* in ciliary flush.
Diclofenac Placebo
N 98 48
Source of variation
Time Period 3-5 Days 7-9 Days 14-16 Days Mean (SE)t Mean (SE) Mean (SE) 0.70 (0.09) 1.44 (0.09) 1.83 (0.09) -0.04 (0.15) -0.08 (0.15) 0.27 (0.18) Significance (P)
Drug <0.001 <0.001 <0.001 Drug <0.001 <0.001 <0.001 controlled for site *Computed as (baseline) - (postoperative). Positive values represent improvements from baseline. t Standard error of the mean
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
Diclofenac Sodium 70,
....
..,
l Placebo
'"
<
' --' --~ L
13'
3"
----
Fig. 4. (Kram Overall assessment of the inflammatory response: Proportion of cases with moderate to marked improvement and those who were the same or worse than baseline at each postoperative period.
73' 80,
14 - 16 Days
7 - 9 Da ys
3 - 5 Days Improvement Over Bassline : •
Moderate or Marked
0
Sl ight
0
Non e
82%
56% 49%
84%
67%
39% ,(._ _______________ _
Placebo Diclofenac 3-5 Days
Fig. 5.
7-9 Days •
DAYS
DAYS
14-16 DAYS
p•0.099
p•0 .001
p•0 .003
14-16 Days •
(Kram Percentage of diclofenac and placebo cases classified as therapeutic failures at each postoperative period. • = P < .001.
tient in the diclofenac sodium group and two in the placebo group were prematurely withdrawn from the study because of complaints of ocular pain. All three cases resolved promptly following discontinuation of the drug or vehicle and oftobramycin. There were no clinically important differences between the groups in the results of slitlamp examinations for safety endpoints at any post-baseline examination.
DISCUSSION For all efficacy variables measured, diclofenac sodium was associated with a significant reduction in inflammation compared with the placebo control consisting of the solution vehicle alone. Diclofenac sodium recipients had significantly better early best cor-
5-7
2- 5
-
Fig. 6.
PLACEBO
[~-:J DICLO
(Kram Percentage of cases in the diclofenac sodium and placebo groups with best corrected visual acuity of 20/40 or better at each postoperative period.
rected visual acuity than placebo recipients, possibly as a result ofless postoperative inflammation and discomfort. There were significantly fewer therapeutic failures in the drug group than in the placebo group. The two groups did not differ significantly in the incidence or severity of adverse reactions. Diclofenac sodium was not associated with any apparent difference in postoperative lOP or in the control of glaucoma in the five glaucomatous patients enrolled. Postoperative lOP rises are common following cataract extraction in patients with glaucoma and are observed in normal patients as well. 7 Steroids used to reduce ocular inflammation following cataract surgery may exacerbate the lOP increase in glaucomatous patients, 8•9 contraindicating their use in these patients. Previous investiga-
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
143
tions of the effects of several NSAIDs on postoperative lOP have found varied responses ranging from inhibition of the lOP increase, 10• 11 to enhanced increase when combined with corticosteroids, 12 to no apparent effect when compared with placebo controls. 13 • 14 Our results were consistent with the latter and suggest an obvious advantage of diclofenac sodium over corticosteroids. This study demonstrated that diclofenac sodium was an effective and well-tolerated NSAlD for patients who have had cataract extraction and IOL implantation. Previous studies have shown that diclofenac sodium performs well when compared with steroidal antiinflammatory drugs without the risk that accompanies steroids.
2-Patient minimally improved compared to baseline 3-Patient the same as baseline 4-Patient worse than baseline REFERENCES l. Ronen S, Rozenman Y, Zylbermann R, Berson D. Treat-
2.
3.
APPENDIX All slitlamp examinations were conducted under standard conditions: dim room illumination, highest lamp voltage, smallest aperture (0.3 mm), illumination angle of 30 degrees, magnification 16x, and observation for five seconds. Anterior Chamber Cells O-None 1-1 to 5 cells 2-6 to 15 cells 3-16 to 30 cells 4-More than 30 cells
5.
6.
Anterior Chamber Rare 0-Absent 1-Trace 2-Mild intensity 3-Moderate intensity 4-Strong intensity
7.
8.
Conjunctival Erythema 0-No erythema 1-Pale reddish color of bulbar conjunctiva involving fine conjunctival vessels 2-Red color of bulbar conjunctiva involving fine conjunctival vessels 3-Bright red color of bulbar conjunctiva involving conjunctival and episcleral vessels 4-Definite chemosis ofbulbar conjunctiva and bulbar conjunctival injection involving conjunctival and episcleral vessels Ciliary Rush 0-No ciliary flush !-Injection covering 2-Injection covering 3-Injection covering 4-Injection covering
up to 25% ofperilimbal region 26% to 50% of perilimbal region 51% to 75% of perilimbal region 76% to 100% of perilimbal region
Overall Assessment of Inflammatory Response 0-Patient markedly improved compared to baseline !-Patient moderately improved compared to baseline
144
4.
9. I 0. 11.
12.
13.
14.
ment of ocular inflammation with diclofenac sodium: double-blind trial following cataract surgery. Ann Ophthalmol 1985; 17:577-581 Ilic J, Gigon S, Leuenberger PM. Comparison de l'effet anti-inflammatoire des collyres de Dexamethasone et de Diclofenac. Klin Monatsbl Augenheilkd 1984; 184:494498 Quentin D, Behrens-Baumann W. Coppelblindstudie uber die Wirksamkeit des Prostaglandinsynthese-Hemmers Diclofenac und Dexamethasonphosphat bei der Behandlung der Iritis nach baler Applikation. Fortschr Ophthalmol 1987; 84:353-355 KrafT MC, Sanders DR, McGuigan L, Raanan MG. Inhibition of blood-aqueous humor barrier breakdown with diclofenac: a fluorophotometric study. Arch Ophthalmol 1990; 108:380-383 Arie M, Sawa M, Takase M. Topical flurbiprofen and diclofenac suppress blood-aqueous barrier breakdown in cataract surgery: a fluorophotometric study. Jpn J Ophthalmol 1983; 27:535-542 Agresti A. Categorical Data Analysis. New York, John Wiley, 1990; 333-336 Savage JA, Thomas JV, Belcher CD, et al. Extracapsular cataract extraction and posterior chamber lens implantation in glaucomatous eyes. Ophthalmology 1985; 92: 1506-1516 Armaly MF. Effect of corticosteroids on intraocular pressure and fluid dynamics. Arch Ophthalmol 1963; 70:482491 Briggs HH. Glaucoma associated with the use of topical corticosteroids. Arch Ophthalmol 1963; 70:312-318 Sanders DR, KraffMC, Lieberman HL, et al. Breakdown and reestablishment of blood-aqueous barrier with implant surgery. Arch Ophthalmol 1982; 100:588-590 Strelow SA, Sherwood MB, Brancato U, et al. The effect of diclofenac sodium ophthalmic solution on intraocular pressure following cataract extraction. Ophthalmic Surg 1992; 23:170-175 Rach AJ, Graham J, Kruger LP, et al. Quantitative assessment of postsurgical breakdown of the blood aqueous barrier following administration of 0.5% ketorolac tromethamine solution. Arch Ophthalmol 1988; 166:344347 Sabiston D, Tessler H, Sumers K, et al. Reduction of inflammation following· cataract surgery by the nonsteroidal anti-inflammatory drug, flurbiprofen. Ophthalmic Surg 1987; 18:873-877 Gormaz A. Ocular tension after cataract surgery. Am J Ophthalmol 1962; 50:832-841 '
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
ABSTRACT One hundred forty-eight patients were enrolled in a randomized, prospective, placebo-controlled clinical trial evaluating the efficacy of diclofenac sodium (Voltaren Ophthalmic®) in reducing ocular inflammation following extracapsular cataract extraction with posterior chamber intraocular lens implantation. Eligible patients were enrolled and randomized (2:1 diclofenac:placebo) if the sum of anterior chamber cells plus flare one day postoperatively (baseline) was at least four. None of the patients received concomitant steroidal anti-inflammatory treatment. The 99 patients receiving diclofenac sodium had significantly greater improvement from baseline in summed flare plus cell score than the 49 placebo patients at two to five days and seven to nine days after baseline. Similarly, diclofenac sodium patients had significantly less post-baseline conjunctival erythema and ciliary flush than placebo patients. Significantly more diclofenac sodium patients than placebo patients showed moderate to marked improvement from baseline in overall assessment of inflammatory response. Forty-nine percent of placebo patients but only 17% of diclofenac patients were considered therapeutic failures (P < .001 ). By five to seven days, 82% of diclofenac sodium patients and 59% of placebo patients had corrected visual acuities of 20/40 or better (P < .001 ). There were no clinically important differences in mean intraocular pressure at any visit.
Key Words: anterior chamber reaction, diclofenac sodium, nonsteroidal antiinflammatory drug, ocular inflammation, post-cataract inflammation
Diclofenac sodium (Voltaren Ophthalmic®) is a potent, rapidly acting nonsteroidal anti-inflammatory drug (NSAID). The oral formulation of diclofenac sodium received Federal Drug Administration (FDA) approval in 1988 for use in treating rheumatoid arthritis, osteoarthritis, and ankylosing spondylitis. Clinical ophthalmologic trials have demonstrated that topical and oral administration of diclofenac sodium effectively reduces a variety of signs and symptoms of ocular inflammation following cataract extraction and intraocular lens (IOL) implantation. A placebo-controlled trial of 100 intracapsular cata-
ract extraction cases by Ronen and coauthors 1 showed significantly decreased levels of anterior chamber flare and cells in the diclofenac treated group compared with a group that did not receive di.clofenac. Both groups received routine postoperative topical steroids. In another study of intracapsular cataract extraction cases, topical diclofenac treatment was compared with dexamethasone treatment; the study demonstrated equivalent efficacy for diclofenac sodium and the steroid. 2 Diclofenac has also been shown to have relatively greater efficacy in reducing ocular inflammation in iritis and iridocyclitis cases than dexamethasone. 3• Other ste-
From Northwestern University and Kraff Eye Institute, Chicago, Illinois (Krajj), Carolina Eye Associates, P.A., and Medical Care International Ophthalmic Institute, Southern Pines, North Carolina (Martin), Neumann Eye Institute, Deland, Florida (Neumann), and Eye Associates of New Mexico, Albuquerque, New Mexico (Weinstein). Reprint requests to Manus C. Kraff, M.D., Kraff Eye Institute, 5600 West Addison Street, Chicago, Illinois 60634. 138
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
ciliary flush (see Appendix for explanation of clinical grades), the overall assessment of inflammatory response (assessing change from baseline; see Appendix), and the therapeutic failure rate. A patient was considered a therapeutic failure if ( 1) he or she was withdrawn from the study because of an adverse experience, (2) if he or she was withdrawn because of an unsatisfactory therapeutic response to treatment, at the discretion of the surgeon, or (3) if the sum of anterior chamber cells and flare was the same or greater than the one-day postoperative (baseline) level. Ophthalmic safety was evaluated by changes from baseline in visual acuity and lOP and by observing change in the slitlamp examination and funduscopy. Instances of ocular or systemic reactions and the need for concomitant prescription medications other than tobramycin were also evaluated. A single clinical examiner performed all assessments at all visits at each site. Two of the examiners were ophthalmologists and two were optometrists. All four clinical examiners were trained together before initiation of the study. Quantitative variables were measured with the same instrument under the same examination conditions for all patients at a site. For postsurgical SUBJECTS AND METHODS inflammatory variables, the examiners from each site Four surgeons at four clinical centers participated in were trained by a lead observer who examined practice this study. Patients of either sex over the age of 35 were patients jointly with the clinical examiners and trained potentially eligible if they were scheduled to have routine them in consistent evaluation. All slitlamp examinacataract extraction by phacoemulsification or the tions were performed under standardized conditions. planned extracapsular method, followed by the implan- The criteria for performing and scoring the slitlamp tation of a posterior chamber IOL. Screening exclusion examination variables are described in detail in the criteria included intraocular pressure (lOP) greater than Appendix. Analysis of efficacy was based on the change in the 24 mm Hg, ocular medications other than hypotensives, of anterior chamber cells and flare from baseline in sum previous surgery in the eye being operated on, history of the treatment and placebo groups. In addition, ratings of ocular inflammation, recent steroid use, allergy to conjunctival erythema, ciliary flush, and overall assessNSAIDs, and blindness or absence of the fellow eye. ment of inflammatory response were compared. In each Potentially eligible patients were screened at one day instance, the groups were compared using a weightedpostoperatively (baseline examination). Patients with a of variance least-squares estimate in a linear analysis sum of anterior chamber cells and flare of four or greater model for mean response, confirmed by Cochran-Man(see Appendix for explanation of clinical grades) were randomized to receive 0.1% diclofenac sodium ophthal- tel-Haenszel tests controlling for potential effects of mic solution or placebo (solution vehicle only). Two study center. The proportions of therapeutic failures in active drug recipients were randomized for each placebo each group were compared with Cochran-MantelHaenszel tests controlling for study center. 6 recipient. If a patient was withdrawn following a study visit, his Patients were instructed to instill one drop of the or her ratings at that visit were carried through for all masked solution four times daily beginning 22 to 34 subsequent visits. All analyses were conducted on an hours postoperatively and continuing for 14 days. Paintent-to-treat basis (see Results). Statistical significance tients were examined at one day postoperatively (baseat P s .05 for all comparisons. All tests were was set line) and at three to five days, seven to nine days, and 14 two-tailed. The sample size was calculated to provide to 16 days postoperatively. One drop of tobramycin ~0.80 for therapeutic failure rate, in the absence power ophthalmic solution 0.3% was administered four times daily on the first day postoperatively and continued at of data to calculate power for efficacy variables. the discretion of the surgeon. Tobramycin was administered at least five minutes after the study solution. RESULTS The primary efficacy endpoint for this study was the A total of 148 patients were enrolled: 99 received change in the sum of anterior chamber cells and flare from baseline. Secondary efficacy endpoints included diclofenac sodium ophthalmic solution and 49 received the clinical assessment of conjunctival erythema and the placebo. Eleven patients (eight diclofenac, three plaraid-controlled studies have established that the preoperative and postoperative topical administration of 0.1% diclofenac sodium solution significantly speeds the re-establishment of the blood-aqueous barrier after surgery. 4 •5 While corticosteroids are not specifically approved by FDA for treating postoperative ocular inflammation, they are widely administered after cataract extraction. Because the efficacy of these steroids in reducing ocular inflammation had not been formally established, the FDA required that the efficacy of new anti-inflammatory drugs be established by placebo-controlled studies. This report presents the results of a multi-site, doublemasked, placebo-controlled clinical trial comparing diclofenac sodium ophthalmic solution with placebo (solution vehicle alone) to determine the anti-inflammatory efficacy, safety, and tolerability of diclofenac sodium following cataract surgery. Topical diclofenac sodium ophthalmic (Voltaren Ophthalmic®) solution received FDA approval for the treatment of post-cataract extraction inflammation in March 1991, based in part on this study.
J CATARACT !WFRACT SURG-VOL 20, MARCH 1994
139
cebo; 7% of the enrollment) were protocol violators with 33 visit violations (21 diclofenac; 12 placebo) of a total of592 planned postoperative visits. Violations included unallowed concomitant medication, excluded preoperative ocular pathology, occasional clinical assessments not performed by independent observers, elevated baseline lOP, and mildly out of range postoperative visits. Because of the low number of protocol violations, and no apparent group differences with respect to protocol violations, all enrolled cases were analyzed on an intentto-treat basis. The patients were predominantly white (95%) and female (61 %). The mean age was 70.6 years (standard deviation 9.3). A total of 26 variables were tested for baseline comparability. Only race differed significantly between the diclofenac (99% white, 1% other) and placebo (88% white, 12% other)groups(whitevs. other, P= .006 by Fisher's exact test), but race was not associated with any treatment response. Most patients had phacoemulsification (74%) rather than extracapsular extraction of their cataracts. At each site, the ratio of phacoemulsification to planned extracapsular extraction was similar in drug and placebo recipients. The method of cataract extraction was not associated with any statistically significant difference in any endpoint examined. Only one patient in each group experienced a posterior capsule rupture and none experienced vitreous loss or posterior capsulotomy. One patient in the drug group experienced a small inferior opening in the capsule and one suffered a partial iris prolapse. No patient had more than 50% hyphema after surgery. Four recipients of diclofenac sodium and 17 recipients of the placebo were prematurely withdrawn from the study. Reasons for withdrawal included adverse reactions (one diclofenac sodium recipient, two placebo recipients), unsatisfactory therapeutic response (three and 12 patients, respectively), and protocol or schedule violations (three placebo recipients). One patient experienced an adverse reaction at the first post-baseline examination prior to diclofenac treatment and was not included in efficacy analyses but was retained in safety analyses. Scores for anterior chamber cells and for anterior chamber flare may range from 0 to 4, and the sum of anterior chamber cells and flare may range from 0 to 8. All patients had a score of 4 or greater at baseline (one day postoperatively), which was the criterion for admission to the study. Tables 1 through 3 present the mean postoperative values for cells, flare, and the sum of cells and flare for patients receiving diclofenac sodium and placebo. Only one comparison, anterior chamber cells at three to five days postoperatively, failed to achieve statistical significance. Figure 1 is a line graph of the mean sum of cells and flare at baseline and at each postoperative time period. While the baseline starting scores were equal in both treatment groups, patients receiving diclofenac sodium 140
Table 1. Change from baseline in mean anterior chamber cell count. Time Period N Diclofenac Placebo
98 49t
3-5 Days 7-9 Days Mean (SE)* Mean (SE)
14-16 Days
1.44 (0.11) 0.80 (0.15)
2.16 (0.10) 1.06 (0.16)
0.70 (0.10) 0.48 (0.14)
Mean (SE)
Significance (P)
Source of variation
0.410 0.186
Drug Drug controlled for site
<0.001 0.001
<0.001 <0.001
* Standard error of the mean
t N = 48 at the 3-5 day visit
Table 2. Change from baseline in anterior chamber flare score. Time Period N Diclofenac Placebo
98 49t
3-5 Days
7-9 Days Mean (SE)* Mean (SE)
14-16 Days
0.49 (0.09) 0.17 (0.11)
1.48 (0.09) 0. 71 (0.14)
Drug controlled for site
Mean (SE)
Significance (P)
Source of variation Drug
0.91 (0.09) 0.35 (0.13)
<0.001 0.026
<0.001 0.001
<0.001 <0.001
* Standard error of the mean
t N = 48 at the 3-5 day visit
had lower mean scores than patients receiving the placebo solution at each time period after baseline. The distribution of summed flare and cell scores shifted from a median value of 7 at baseline to 5 at three to five days, to 4 at seven to nine days, and to 2 at 14 to 16 days in the diclofenac group. Patients receiving diclofenac sodium had greater improvement from baseline in summed flare and cell score than patients receiving placebo at all three post-baseline examinations (Table 3). Significant site differences were noted in the scores for anterior chamber cells, anterior chamber flare, and the sum of flare and cells. There was only one statistically significant interaction between drug and site, for anterior chamber flare at three to five days postoperatively (P :S .02). The differences between drug and placebo groups were statistically significant at all sites, evaluated individually, and the overall difference between drug and placebo groups remained statistically significant when adjusted for the differences among sites.
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
Table 3. Change from baseline in sum of cells and flare. Time Period 7-9 Days 3-5 Days Mean (SE)* Mean (SE) 1.19 (0.17)t 2.35 (0.18)
N
Diclofenac Placebo
98 48
1.14 (0.25)
0.65 (0.22)
14-16 Days Mean (SE) 3.64 (0.17) 1.78 (0.27)
Significance (P)
Source of variation
<0.001 <0.001 Drug 0.016 <0.001 <0.001 Drug 0.052 controlled for site * Standard error of the mean t Computed as (baseline sum of flare and cells) - (postoperative sum of flare and cells). Positive values represent improvements from baseline.
Figures 2 and 3 are graphs of mean conjunctival erythema and mean ciliary flush versus time for patients receiving diclofenac sodium or placebo. For both measures, the means were significantly lower for patients receiving diclofenac sodium than for those receiving the placebo at all post-baseline examinations. For all comparisons, the effect of diclofenac sodium, controlled for study site, was significant at P s .001 (Tables 4 and 5). Overall assessment of inflammatory response was graded as moderate to marked improvement over baseline or the same or worse than baseline. Figure 4 presents the proportions of cases with "moderate to marked improvement" and "same or worse than baseline" by treatment group for each postoperative period. Significantly more recipients of diclofenac sodium than recipients of placebo displayed moderate to marked improvement at seven to nine days and at 14 to 16 days relative to baseline (P ::5 .001 ). Significantly more recipients of placebo than of diclofenac sodium were rated as being
the same or worse than at baseline (P s .01 at all three periods). The rates of therapeutic failures are presented in Figure 5. At seven to nine days postoperatively, 49% of placebo recipients and 17% of diclofenac sodium recipients were considered therapeutic failures (P ::s; .001 ). At 14 to 16 days postoperatively, 39% of placebo recipients but only 4% of diclofenac sodium recipients were therapeutic failures (P ::5 .001). At two to five days postoperatively, 67% of diclofenac sodium recipients had best corrected visual acuity of 20/40 or better compared with 53% of placebo recipients (Figure 6). By five to seven days, the difference (82% diclofenac vs. 56% placebo) was statistically significant (P s .001), and it remained significant at 14 to 16 days postoperatively (84% vs. 59%, P s .001). At the final study (endpoint) visit, 80% of diclofenac sodium recipients had a two-line improvement in best corrected vision relative to their presurgery baseline, while only 57% of placebo recipients had a two-line improvement (P ::s; .0 1). The possible confounding effect of phacoemulsification versus planned extracapsular extraction on visual acuity was examined. Extraction technique was not associated with a statistically significant difference in visual acuity, nor was a statistically significant interaction with drug treatment found. There were no clinically important differences between the groups in the frequency or severity of postoperative changes in lOP at any visit. Mean endpoint lOP was 16.6 mm Hg for the diclofenac group and 15.8 mm Hg for the placebo group. Mean change from baseline lOP after two to five treatment days was -2.7 mm Hg for diclofenac and -2.8 mm Hg for placebo. Four glaucoma patients received diclofenac sodium and one received placebo. There were no adverse effects on glaucoma control. There were no significant differences between the groups in the frequency of adverse outcomes. One pa-
8 F 7 L A 6
R E
5
+
4
c
3
E L L
Fig. 1. **
2
s
0 0
2
3
4
5
6
8
7
9
10
11
Days On Treatment ----- Diclofenac
~ Placebo
12
13
14
15
(Kraft) Mean sum of cells and flare at baseline and each postoperative period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. * = P ~ .052; ** = P < .05.
16
I
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
141
c 0
4~-------------------
n j
Fig. 2.
(Kraft) Mean conjunctival erythema score at baseline and each postoperative period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. Erythema: 0 = none; 1 = pale red; 2 = red; 3 = bright red; 4 = definite chemosis. * = p < .01.
u n t i v
3
a
2
I
.----------~~
E
r y t h
e
oL-~~-L~--L-~~~~--L-~~~~--L-~-L~~~
m
0
a
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20
I
Days On Treatment
I ___,._ Diclofenac
--+-- Placebo
4~----------------------------------------------,
c I L I
Fig. 3.
3
(Kraft) Mean ciliary flush score at ~ baseline and each postoperative y period for diclofenac sodium and placebo patients. Vertical lines represent ± standard error. Per- u centage of perilimbal region cov- S ered: 0 =none; 1 = 1%-25%; 2 = H 26%-50%; 3 = 51%-75%; 4 = 76%-100%. * = p < .01.
•
2
E
oL_~~_L~--L_~-L~~--L_J__L_J
0
1
2
3
4
5
6
7
8
_ _L_J__L~~L_~~
9 10 11 12 13 14 15 16 17 18 19 20
Days On Treatment ___,._ Diclofenac
Table 4. Mean difference* in conjunctival erythema.
Diclofenac Placebo Source of variation
N 98 48
Time Period 3-5 Days 7-9 Days Mean (SE)t Mean (SE) 0.66 (0.08) 1.38 (0.09) -0.08 (0.13) 0.02 (0.15)
14-16 Days Mean (SE) 1.73 (0.09) 0.39 (0.16)
Significance (P)
Drug <0.001 <0.001 <0.001 Drug <0.001 <0.001 <0.001 controlled for site *Computed as (baseline) - (postoperative). Positive values represent improvements from baseline. t Standard error of the mean
142
--+-- Placebo
I
Table 5. Mean difference* in ciliary flush.
Diclofenac Placebo
N 98 48
Source of variation
Time Period 3-5 Days 7-9 Days 14-16 Days Mean (SE)t Mean (SE) Mean (SE) 0.70 (0.09) 1.44 (0.09) 1.83 (0.09) -0.04 (0.15) -0.08 (0.15) 0.27 (0.18) Significance (P)
Drug <0.001 <0.001 <0.001 Drug <0.001 <0.001 <0.001 controlled for site *Computed as (baseline) - (postoperative). Positive values represent improvements from baseline. t Standard error of the mean
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
Diclofenac Sodium 70,
....
..,
l Placebo
'"
<
' --' --~ L
13'
3"
----
Fig. 4. (Kram Overall assessment of the inflammatory response: Proportion of cases with moderate to marked improvement and those who were the same or worse than baseline at each postoperative period.
73' 80,
14 - 16 Days
7 - 9 Da ys
3 - 5 Days Improvement Over Bassline : •
Moderate or Marked
0
Sl ight
0
Non e
82%
56% 49%
84%
67%
39% ,(._ _______________ _
Placebo Diclofenac 3-5 Days
Fig. 5.
7-9 Days •
DAYS
DAYS
14-16 DAYS
p•0.099
p•0 .001
p•0 .003
14-16 Days •
(Kram Percentage of diclofenac and placebo cases classified as therapeutic failures at each postoperative period. • = P < .001.
tient in the diclofenac sodium group and two in the placebo group were prematurely withdrawn from the study because of complaints of ocular pain. All three cases resolved promptly following discontinuation of the drug or vehicle and oftobramycin. There were no clinically important differences between the groups in the results of slitlamp examinations for safety endpoints at any post-baseline examination.
DISCUSSION For all efficacy variables measured, diclofenac sodium was associated with a significant reduction in inflammation compared with the placebo control consisting of the solution vehicle alone. Diclofenac sodium recipients had significantly better early best cor-
5-7
2- 5
-
Fig. 6.
PLACEBO
[~-:J DICLO
(Kram Percentage of cases in the diclofenac sodium and placebo groups with best corrected visual acuity of 20/40 or better at each postoperative period.
rected visual acuity than placebo recipients, possibly as a result ofless postoperative inflammation and discomfort. There were significantly fewer therapeutic failures in the drug group than in the placebo group. The two groups did not differ significantly in the incidence or severity of adverse reactions. Diclofenac sodium was not associated with any apparent difference in postoperative lOP or in the control of glaucoma in the five glaucomatous patients enrolled. Postoperative lOP rises are common following cataract extraction in patients with glaucoma and are observed in normal patients as well. 7 Steroids used to reduce ocular inflammation following cataract surgery may exacerbate the lOP increase in glaucomatous patients, 8•9 contraindicating their use in these patients. Previous investiga-
J CATARACT REFRACT SURG-VOL 20, MARCH 1994
143
tions of the effects of several NSAIDs on postoperative lOP have found varied responses ranging from inhibition of the lOP increase, 10• 11 to enhanced increase when combined with corticosteroids, 12 to no apparent effect when compared with placebo controls. 13 • 14 Our results were consistent with the latter and suggest an obvious advantage of diclofenac sodium over corticosteroids. This study demonstrated that diclofenac sodium was an effective and well-tolerated NSAlD for patients who have had cataract extraction and IOL implantation. Previous studies have shown that diclofenac sodium performs well when compared with steroidal antiinflammatory drugs without the risk that accompanies steroids.
2-Patient minimally improved compared to baseline 3-Patient the same as baseline 4-Patient worse than baseline REFERENCES l. Ronen S, Rozenman Y, Zylbermann R, Berson D. Treat-
2.
3.
APPENDIX All slitlamp examinations were conducted under standard conditions: dim room illumination, highest lamp voltage, smallest aperture (0.3 mm), illumination angle of 30 degrees, magnification 16x, and observation for five seconds. Anterior Chamber Cells O-None 1-1 to 5 cells 2-6 to 15 cells 3-16 to 30 cells 4-More than 30 cells
5.
6.
Anterior Chamber Rare 0-Absent 1-Trace 2-Mild intensity 3-Moderate intensity 4-Strong intensity
7.
8.
Conjunctival Erythema 0-No erythema 1-Pale reddish color of bulbar conjunctiva involving fine conjunctival vessels 2-Red color of bulbar conjunctiva involving fine conjunctival vessels 3-Bright red color of bulbar conjunctiva involving conjunctival and episcleral vessels 4-Definite chemosis ofbulbar conjunctiva and bulbar conjunctival injection involving conjunctival and episcleral vessels Ciliary Rush 0-No ciliary flush !-Injection covering 2-Injection covering 3-Injection covering 4-Injection covering
up to 25% ofperilimbal region 26% to 50% of perilimbal region 51% to 75% of perilimbal region 76% to 100% of perilimbal region
Overall Assessment of Inflammatory Response 0-Patient markedly improved compared to baseline !-Patient moderately improved compared to baseline
144
4.
9. I 0. 11.
12.
13.
14.
ment of ocular inflammation with diclofenac sodium: double-blind trial following cataract surgery. Ann Ophthalmol 1985; 17:577-581 Ilic J, Gigon S, Leuenberger PM. Comparison de l'effet anti-inflammatoire des collyres de Dexamethasone et de Diclofenac. Klin Monatsbl Augenheilkd 1984; 184:494498 Quentin D, Behrens-Baumann W. Coppelblindstudie uber die Wirksamkeit des Prostaglandinsynthese-Hemmers Diclofenac und Dexamethasonphosphat bei der Behandlung der Iritis nach baler Applikation. Fortschr Ophthalmol 1987; 84:353-355 KrafT MC, Sanders DR, McGuigan L, Raanan MG. Inhibition of blood-aqueous humor barrier breakdown with diclofenac: a fluorophotometric study. Arch Ophthalmol 1990; 108:380-383 Arie M, Sawa M, Takase M. Topical flurbiprofen and diclofenac suppress blood-aqueous barrier breakdown in cataract surgery: a fluorophotometric study. Jpn J Ophthalmol 1983; 27:535-542 Agresti A. Categorical Data Analysis. New York, John Wiley, 1990; 333-336 Savage JA, Thomas JV, Belcher CD, et al. Extracapsular cataract extraction and posterior chamber lens implantation in glaucomatous eyes. Ophthalmology 1985; 92: 1506-1516 Armaly MF. Effect of corticosteroids on intraocular pressure and fluid dynamics. Arch Ophthalmol 1963; 70:482491 Briggs HH. Glaucoma associated with the use of topical corticosteroids. Arch Ophthalmol 1963; 70:312-318 Sanders DR, KraffMC, Lieberman HL, et al. Breakdown and reestablishment of blood-aqueous barrier with implant surgery. Arch Ophthalmol 1982; 100:588-590 Strelow SA, Sherwood MB, Brancato U, et al. The effect of diclofenac sodium ophthalmic solution on intraocular pressure following cataract extraction. Ophthalmic Surg 1992; 23:170-175 Rach AJ, Graham J, Kruger LP, et al. Quantitative assessment of postsurgical breakdown of the blood aqueous barrier following administration of 0.5% ketorolac tromethamine solution. Arch Ophthalmol 1988; 166:344347 Sabiston D, Tessler H, Sumers K, et al. Reduction of inflammation following· cataract surgery by the nonsteroidal anti-inflammatory drug, flurbiprofen. Ophthalmic Surg 1987; 18:873-877 Gormaz A. Ocular tension after cataract surgery. Am J Ophthalmol 1962; 50:832-841 '
J CATARACT REFRACT SURG-VOL 20, MARCH 1994