ABSTRACT
Purpose: To evaluate the effectiveness of diclofenac eyedrops in reducing inflamma-
tion and the incidence of angiographic cystoid macular edema (CME) after cataract
surgery and intraocular lens (IOL) implantation.
Setting: Eye Clinic, Institute of Biomedical Sciences, San Paolo Hospital, Milan, Italy.
.
anglOgrap h'lC CME WIt . h IO - 12 or Wit
• h out 13 t h e concur- measurement by applanation tonometry, and ophthal-
rent use of steroids. moscopic evaluation.
Diclofenac sodium (Voltaren®) is a potent, rapidly The number of anterior chamber cells and flare were
acting NSAID, recently approved by the U.S. Food and determined by two well-trained, masked evaluators us-
Drug Administration (FDA) for the treatment of post- ing the following clinical grades. For anterior chamber
cataract extraction inflammation. Although steroids are cells, 0 = none, 1 = 1 to 5 cells, 2 = 6 to 15 cells,
not specifically approved by FDA for the treatment of 3 = 16 to 30 cells, 4 = more than 30 cells; for flare,
postoperative ocular inflammation, they are widely ad- o = none, 1 = trace of flare, 2 = mild flare, 3 = mod-
ministered as standard therapy after cataract extraction. erate flare, 4 = strong flare. All the evaluations were
This randomized clinical trial was designed to evaluate made under standard conditions: dim room illumina-
whether diclofenac sodium ophthalmic solution should tion, the highest lamp voltage, the smallest aperture
be added to the standard therapy to reduce postoperative (0.3 mm), a 30 degree angle, and a magnification of
inflammation and the incidence of angiographic CME 16X.
after cataract surgery. Contrast sensitivity was assessed 1, 3, and 6 months
after surgery using a computer-based system (Conel Sys-
tem) that presents sinusoidal bars on a screen and mea-
Subjects and Methods sures contrast sensitivity at five levels of spatial
Patients scheduled for extracapsular cataract extrac- frequency. Contrast sensitivity was evaluated with best
tion (ECCE) with implantation of an intraocular lens corrected refraction.
(IOL) were eligible for inclusion in the trial. We ex- All patients were scheduled for fluorescein angiog-
cluded patients with diabetes, glaucoma, and a diagnosis raphyat 1, 3, and 6 months postoperatively. The angio-
of maculopathy, and those being systemically treated grams were read and graded according to the Miyake
with steroids, acetazolamide, or NSAIDs. After giving classification 11 by one masked retinal specialist: 0 = no
informed consent, the patients were randomized to re- leakage, 1 = less than perifovealleakage, 2 = clinically/
ceive 0.1 % diclofenac sodium ophthalmic solution or a ophthalmoscopically evident 360 degree perifovealleak-
placebo. Randomization was obtained using a table of age, and 3 = severe (>one disc diameter).
random numbers. The trial was double-masked. The A two-sample t-test was used to test for significance
eyedrops were administered by the nursing staff four of differences in group means. The chi-square test, Fish-
times daily starting 3 days before surgery until patient er's exact test, and trend analysis were used for statistical
discharge. Patients were then instructed to instill one analysis of proportions.
drop four times daily for 3 months after surgery.
The majority of the surgeries were performed by one
surgeon (N.O.) using both conventional ECCE and
phacoemulsification. Approximately half the ECCEs
Results
were performed by residents in training. Eighty-eight patients were included in the study:
All patients received a 2 ml sub-Tenon' s injection of 42 received diclofenac eyedrops and 46, placebo. In the
betamethasone (1 m!) and amikacin (1 m!) at the end of diclofenac group, 14 eyes (33.3%) had phacoemulsifica-
surgery and 1 day postoperatively. Drops containing to- tion and 28 (66.7%), ECCE. In the placebo group, 11
bramycin and dexamethasone were administered four (23.9%) had phacoemulsification and 35 (76.1%),
times daily for 3 weeks after surgery. In addition, a myd- ECCE. All patients completed the study, and treatment
riatic (tropicamide 1%) was instilled twice daily during regimens were well tolerated with no evidence of rele-
the first postoperative week. vant side effects.
Complete ocular examinations were performed and Patient characteristics are shown in Table 1. There
recorded 2 days before surgery, at 1, 5, and 28 days were no significant differences between the two groups
postoperatively, and at 3 and 6 months. Examinations in age, sex, and preoperative visual acuity. Complica-
included visual acuity assessment (Snellen chart), slit- tions and additional surgical procedures are shown in
lamp biomicroscopy, intraocular pressure (lOP) Table 2. The posterior capsule ruptured in eight pa-
was associated with less postoperative inflammation Postoperative Mean (SO) Mean (SO)
Month >20/40 (%) >20/40 (%)
than placebo. The difference between the two groups
was statistically significant at 5 days (Tables 3 and 4). 0.76 (0.26) 83 0.72 (0.26) 80
3 0.84 (0.22) 88 0.83 (0.27) 89
Intraocular pressure was similar in the two groups at
6 0.90 (0.25) 93 0.90 (0.24) 91
all follow-up visits (P = .38).
Visual acuity at 1, 3, and 6 months postoperatively There was no significant difference in mean visual acuity between
the groups
is shown in Table 5. There were no significant differ-
ences between the two groups. Mean contrast sensitivity
was higher in the diclofenac patients (Figure 1), al- (Table 6). Of the eight cases, one (in the placebo group)
though the difference between the two groups was sta- had clinical signs of CME and was given 250 mg of
tistically significant only at 1 month and for higher acetazolamide orally twice daily. The other seven cases
spatial frequencies (10.5 cycles per degree). had only angiographic evidence, with normal postoper-
At 1 month postoperatively, eight patients (9.0%) ative visual acuities. However, five had significantly
had angiographic evidence of CME: seven in the placebo lower contrast sensitivities. Five of the eight cases were
group and one in the diclofenac group (P = .039). At self-limiting by 6 months postoperatively and were as-
3 and 6 months, this difference was no longer significant sociated with an intraoperative complication or an ad-
120,---------------------------------------------------~
Table 6. Cases of angiographic CME. day, the anti-inflammatory effect was greater and the
Diclofenac Placebo diclofenac-treated eyes had significantly less flare and
Group Group fewer anterior chamber cells than the placebo-treated
Postoperative Month Number (%) Number (%) eyes.
l' (2.4) 7 (15.2) Diclofenac treatment was not associated with in-
3 (2.4) 4 (8.7) creased postoperative lOP. Some previous studies have
6 (2.4) 2 (4.3)
concluded that NSAID treatment can influence postop-
With complicated surgery 1/3 (33.3) 4/5 (80.0)
erative lOP, particularly when NSAIDs are combined
'Chi-square = 4.38 (P = .04) . h sterol'ds. 16-18 In th e current study, t he group
wIt
treated with the combination of diclofenac and steroids
ditional surgical procedure. No CME occurred later had lOP values very similar to those observed in the
than 1 month postoperatively. group receiving the combination of placebo and ste-
roids, suggesting that diclofenac did not affect lOP.
Although visual acuities in the two groups were sim-
Discussion ilar, the diclofenac-treated patients had better contrast
sensitiviry; the difference between them and the place-
In this study, the diclofenac-treated eyes had less
postoperative inflammation and a lower incidence of bo-treated patients was particularly evident for higher
angiographic CME than the placebo-treated eyes. Pre- spatial frequencies. This finding confirms Ginsburg and
vious reports have shown that diclofenac sodium effec- coauthors'19 findings for other NSAIDs. Flurbiprofen
tively decreases ocular inflammation after cataract treatment improved contrast sensitivity in patients with
surgery and 10L implantation. 5,6,9,14,15 The results of and without CME significantly at 12 cycles per degree
this study suggest that diclofenac can enhance the effect on postoperative day 40. In our study, the higher con-
of steroids in reducing postoperative inflammation. trast sensitivities observed in the diclofenac-treated pa-
This enhancement is probably because the anti-inflam- tients did not seem to be due to the reduced
matory effect of the two drugs is obtained by two sepa- postoperative inflammation in the anterior chamber
rate mechanisms. Diclofenac seemed to become since 1 month after surgery, there was no inflammation
effective 24 hours after surgery, probably due to the in either group. Perhaps other mechanisms are involved
preoperative administration. On the fifth postoperative in the positive effect of NSAIDs on contrast sensitivity.
Diclofenac treatment was effective in reducing the with diclofenac; a fluorophotometric study. Arch Oph-
incidence of angiographic CME, which supports the thalmol1990; 108:380-383
6. Araie M, Sawa M, Takase M. Topical flurbiprofen and
conclusions of several investigators about the effective-
didofenac suppress blood-aqueous barrier breakdown in
ness of a variety of NSAIDs in preventing angiographic
cataract surgery: a fluorophotometric study. Jpn J Oph-
CME. 10 - 13 One month after surgety the incidence of thalmol1983; 27:535-542.
angiographic CME was approximately 9.0%, which is 7. Flach AJ, Jaffe NS, Akers WA. The effect of ketorolac
consistent with other published data. 10-13 In our study, tromethamine in reducing postoperative inflammation:
as in a previous one,20 the majorityofCME cases were in double-mask, parallel comparison with dexamethasone.
Ann Ophthalmol1989; 21:407-411
eyes in which surgical complications had occurred. Al-
8. Flach AJ, Kraff MC, Sanders DR, Tanenbaum L. The
though "clinical" CME is today an infrequent compli-
quantitative effect of 0.5% ketorolac tromethamine solu-
cation21 following cataract surgery (perhaps less than tion and 0.1 % dexamethasone sodium phosphate solu-
1.0% with advanced surgical procedures), preventing tion on postsurgical blood-aqueous barrier. Arch
the condition is important because of the number of Ophthalmol 1988; 106:480 - 483
cataract surgeries performed every year (about 1 million 9. Alpar JJ, McGuigan L, Alpar AJ, et al. Effect of didofenac
in the United States alone). A recent meta-analysis of sodium ophthalmic 1% on inflammation following cat-
aract extraction-a fluorophotometric study (abstract).
published literature on preventing CME indicated that
Ophthalmology 1989; 96(suppl): 109
NSAID treatment can decrease the incidence of clini- 10. KraffMC, Sanders DR, Jampol LM, et al. Prophylaxis of
cally relevant CME and improve vision. 22 pseudophakic cystoid macular edema with topical indo-
The primary objective of our study was to assess the methacin. Ophthalmology 1982; 89:885-890
effectiveness of one NSAID when added to the standard 11. Miyake K, Sakamura S, Miura H. Long-term follow-up
management of cataract surgery. The results suggest that study on the prevention of aphakic cystoid macular oe-
dema by topical indomethacin. Br J Ophthalmol 1980;
there are several advantages to using diclofenac in pa-
64:324-328
tients who have had cataract extraction and IOL implan- 12. Yannuzzi LA, Landau AN, Turtz AI. Incidence of apha-
tation. This study, of course, has the drawbacks of small kic cystoid macular edema with the use of topical indo-
trials, and its results must therefore be interpreted cau- methacin. Ophthalmology 1981; 88:947-954
tiously. Future studies should investigate whether 13. Flach AJ, Stegman RC, Graham J, Kruger LP. Prophy-
NSAIDs could completely replace the use of steroids in laxis of aphakic cystoid macular edema without cortico-
steroids; a paired-comparison, placebo-controlled
the management of cataract surgery.
double-masked study. Ophthalmology 1990; 97:1253-
1258
14. Ronen S, Rozenman Y, Zylbermann R, Berson D. Treat-
References ment of ocular inflammation with diclofenac sodium:
1. Flach AJ. Cydo-oxygenase inhibitors in ophthalmology. double-blind trial following cataract surgery. Ann Oph-
Surv OphthalmoI1992; 36:259-284 thalmol1985; 17:577-581
2. Kraff MC, Martin RG, Neumann AC, Weinstein AJ. 15. Hic J, Gigon S, Leuenberger PM. Comparaison de l'effet
Efficacy of diclofenac sodium ophthalmic solution versus anti-inflammatoire des collyres de Dexamethasone et de
placebo in reducing inflammation following cataract ex- Diclofenac. Klin Monatsbl Augenheilkd 1984;
traction and posterior chamber lens implantation. J Cat- 184:494-498
aract Refract Surg 1994; 20:138-144 16. Strelow SA, Driebe WT Jr, Sherwood MB, et al. The
3. Flach AJ, Lavelle q, Olander KW, et al. The effect of effect of didofenac sodium ophthalmic solution on in-
ketorolac tromethamine solution 0.5% in reducing post- traocular pressure following cataract extraction. Oph-
operative inflammation after cataract extraction and in- thalmic Surg 1992; 23:170-175
traocular lens implantation. Ophthalmology 1988; 95: 17. Sanders DR, KraffMC, Lieberman HL, et al. Breakdown
1279-1284 and reestablishment of blood-aqueous barrier with im-
4. Sanders DR, Kraff M. Steroidal and nonsteroidal anti- plant surgery. Arch Ophthalmol1982; 100:588-590
inflammatory agents; effect on postsurgical inflammation 18. Sabiston 0, Tessler H, Sumers K, et al. Reduction of
and blood-aqueous human barrier breakdown. Arch inflammation following cataract surgery by the nonste-
Ophthalmol1984; 102:1453-1456 roidal anti-inflammatory drug, flurbiprofen. Ophthal-
5. KraffMC, Sanders DR, McGuigan L, Raanan MG. In- mic Surg 1987; 18:873-877
hibition of blood-aqueous humor barrier breakdown 19. Ginsburg AP, Cheetham JK, DeGryse RE, Abelson M.
The effects of flurbiprofen and indomethacin on acute literature on visual acuity and complications following
cystoid macular edema: functional vision and contrast cataract extraction with intraocular lens implanta-
sensitivity. ARVO abstract 1153. Invest Ophthalmol Vis tion. Arch Ophthalmol1994; 112:239-252; correction
Sci 1994; 35:1504 p 889
20. Yannuzzi LA. A perspective on the treatment of aphakic 22. Rossetti L, Chaudhuri J, Dickersin K. Effectiveness of
cystoid macular edema. Surv Ophthalmol 1984; treatment to prevent and treat cystoid macular edema: a
28:540-553 meta-analysis. ARVO abstract 2556. Invest Ophthalmol
21. Powe NR, Schein OD, Gieser Sc, et al. Synthesis of the Vis Sci 1993; 34:1224