Worldwide, cataract is the leading cause of preventable and treatable blindness.1 Cataract surgery can improve visual acuity in many patients with mild to severe visual impairment and is considered one of the most cost-effective of all healthcare interventions.2 Even though improvements in modern cataract surgery techniques have significantly decreased the incidence of postoperative complications, cystoid macular edema (CME) remains one of the most important causes of suboptimum visual acuity after otherwise uneventful surgery.3,4
Cystoid macular edema after cataract surgery, also known as the Irvine-Gass syndrome, was first reported by Irvine in 19535 and discussed by Gass and Norton in 1966.6 Since then, ophthalmologists have aimed to prevent CME using perioperative topical corticosteroids and/or nonsteroidal antiinflammatory drugs (NSAIDs), albeit current practice varies between organizations and countries.3,7 As yet, the optimum approach in routine cataract surgery cases remains debatable because of the high societal costs involved with routine prescription of topical NSAIDs in some countries, such as the United States of America.8 An Ophthalmic Technology Assessment of the American Academy of Ophthalmology3 does not support the use of topical NSAIDs because of insufficient high-quality evidence with regard to its long-term benefit to prevent vision loss from CME 3 months or more after cataract surgery. The Cataract Clinical Committee of the American Society of Cataract and Refractive Surgery, on the other hand, recently published a report on topical NSAIDs as an important adjunctive tool for surgeons performing routine and complicated cataract surgery and emphasized the compelling effectiveness of NSAIDs to reduce pain, prevent intraoperative miosis, modulate postoperative inflammation, and reduce the incidence of CME.7 Two recent systematic reviews of the Cochrane Collaboration and 2 independent European metaanalyses suggest that topical NSAIDs reduce the risk for developing CME after cataract surgery, although none of them found a clinically relevant effect on mean visual acuity.9–12
To further investigate this important clinical question, many ophthalmologists have studied the efficacy of various NSAIDs and corticosteroids to prevent CME after cataract surgery. Most studies compared the efficacy of topical NSAIDs or combination treatment versus corticosteroids alone; however, few studies directly compared the efficacy of combination treatment versus topical NSAIDs.9 Moreover, the low a priori incidence of CME demands large samples to detect small, yet clinically important, differences.
The PREvention of Macular EDema after cataract surgery (PREMED) study was a randomized controlled clinical multicenter trial designed to directly compare the efficacy of a topical NSAID, a topical corticosteroid, and the combination of both in nondiabetic patients. The study was funded by the European Society of Cataract and Refractive Surgeons (ESCRS). The aim of the ESCRS PREMED study was to provide evidence-based recommendations that could serve as a basis for clinical guidelines on the prevention of CME after cataract surgery.
Patients and methods
The ESCRS PREMED study was a randomized controlled trial (RCT). The study protocol was approved by the local ethics committees and national authorities of all participating study centers. The study procedures were performed in accordance with the tenets of the Declaration of Helsinki. The study protocol can be found on the U.S. National Institutes of Health Clinical Trials site.A A data safety monitoring board (members reported in Appendix S1, available at http://jcrsjournal.org) evaluated the safety of trial participants.
Patient recruitment started in Maastricht, the Netherlands, in July 2013. Final inclusion took place at study centers in the Netherlands after a decision was made to stop recruitment on February 8, 2016. The trial included nondiabetic patients 21 years or older who required regular phacoemulsification cataract surgery in at least 1 eye. Patients were included in 1 of 12 study centers involved in the ESCRS PREMED study located in Austria (Hospital of the Brothers of St. John of God and Vienna Institute for Research in Ocular Surgery, Vienna), Belgium (Antwerp University Hospital, Antwerp), Germany (Goethe University Hospital, Frankfurt am Main), Portugal (Centro Hospitalar Universitário Coimbra, Coimbra), Spain (Institute of Ocular Microsurgery, Barcelona), and the Netherlands (VU University Medical Center, Amsterdam; Zuyderland Medical Center, Heerlen; Eye Hospital Zonnestraal, Hilversum; University Eye Clinic Maastricht UMC, Maastricht; Haaglanden Medical Center, The Hague; and Elisabeth-TweeSteden Hospital, Tilburg). All patients signed written informed consent before inclusion.
Only 1 eye per patient was included in the study, and patients were excluded if they had sustained moderate to severe visual impairment in the other eye according to the definition of the International Statistical Classification of Diseases and Related Health Problems 10th revision.13 Patients were excluded if they had previous CME, any macular pathology that could influence visual acuity, previous intraocular inflammation or uveitis, retinal vein occlusion, posttraumatic cataract, progressive glaucoma, intraocular pressure (IOP) of 25 mm Hg or higher, previous steroid-induced IOP elevation, pseudoexfoliation syndrome, or Fuchs endothelial dystrophy in the study eye. Furthermore, patients were excluded if they had intraocular surgery in the study eye. Patients who used topical NSAIDs, corticosteroids, antiglaucomatous medication, or high-dose systemic corticosteroids at the time of screening were excluded, as were patients who received an intravitreal injection of bevacizumab or ranibizumab in the study eye in the previous 6 weeks, an intravitreal injection of aflibercept in the previous 10 weeks, or an intraocular or periocular corticosteroid injection in the previous 4 months. Finally, patients were excluded if there was a contraindication to the use of any investigation drug.
Surgical Technique, Patient Allocation, and Treatment
All patients had regular phacoemulsification cataract surgery with intraocular lens (IOL) implantation in the posterior segment and received perioperative and/or postoperative antibiotics according to the standard of care in the participating study center.
Patients were randomly allocated to 1 of 3 treatment groups in a 1:1:1 ratio. Stratified block randomization was performed per study center by a local investigator using concealed online softwareB (ALEA, version 3.0, Formsvision BV) and a block size of 15 patients. Trial participants were unblinded for the allocated treatment. Patients in the bromfenac group received bromfenac 0.09% eyedrops (Yellox) twice daily for 2 days preoperatively and 2 weeks postoperatively. Patients in the dexamethasone group received dexamethasone disodium phosphate 0.1% eyedrops 4 times daily for 2 days preoperatively and 1 week postoperatively, with 1 drop less per day every following week. Patients in the combination treatment group received topical bromfenac and dexamethasone in the above-mentioned doses. Topical treatments started 2 days before cataract surgery because previous studies14,15 found a lower incidence of CME in patients who started antiinflammatory treatment preoperatively. No other ocular corticosteroids or NSAIDs were allowed to be used during the course of the study.
An extensive ophthalmologic examination, including slitlamp evaluation, IOP measurement, and fundoscopy of both eyes, was performed at baseline. Cataract was graded according to the Lens Opacities Classification System II,16 and the presence of aqueous cells and flare was graded according to the Standardization of Uveitis Nomenclature (SUN) classification.17 Subjective refraction and corrected distance visual acuity (CDVA) were measured using Early Treatment Diabetic Retinopathy Study (ETDRS) charts. If patients were unable to read any letter on the ETDRS chart, hand motion or finger counting at a given distance were converted to the logarithm of the minimum angle of resolution (logMAR) equivalent.18
A baseline optical coherence tomography (OCT) was performed within 3 months preoperatively using spectral-domain OCT (SD-OCT) devices (3D OCT, Topcon Corp.; Cirrus HD-OCT, Carl Zeiss Meditec AG; OCT-HS100, Canon, Inc.; RTVue-100 Optovue, Inc.; or Spectralis, Heidelberg Engineering, Inc.). Each study center used only 1 type of SD-OCT device for all preoperative and postoperative measurements. Retinal thickness was measured in accordance with the ETDRS retinal thickness map (Figure 1), which includes the central subfield mean macular thickness in the central 1.0 mm area, parafoveal thickness in a concentric ring located 1.0 to 3.0 mm around the fovea, perifoveal thickness in a concentric ring located 3.0 to 6.0 mm around the fovea, and total macular volume in the central 6.0 mm area.19 If the preoperative OCT was of insufficient quality (eg, because of dense cataract), an alternative baseline measurement was performed at 1 day postoperatively because previous studies have shown no significant differences between preoperative and 1-day postoperative retinal thickness measurements.20
Postoperative visits were performed 6 weeks and 12 weeks postoperatively and included a full ophthalmologic examination of the study eye, as reported above. Postoperative CDVA measurements and SD-OCT assessments were performed by a local investigator who was masked to the allocated study treatment.
Cystoid macular edema usually occurs within 3 months postoperatively, with a peak incidence 4 to 6 weeks after cataract surgery.21 Therefore, the primary outcome of this study was the difference in central subfield mean macular thickness 6 weeks postoperatively compared with the baseline value. Secondary outcomes were the difference in central subfield mean macular thickness 12 weeks postoperatively, the incidence of CME and clinically significant macular edema (CSME) within 6 to 12 weeks postoperatively, and CDVA, total macular volume, and parafoveal and perifoveal thickness 6 weeks and 12 weeks postoperatively. Furthermore, all adverse events were reported.
Cystoid macular edema was defined as an increase in central subfield mean macular thickness of 10% or more over baseline, with cystic changes on SD-OCT. Cystoid changes and other retinal pathology were identified by 2 independent and masked retina specialists of the University Eye Clinic Maastricht UMC (Appendix S1, available at http://jcrsjournal.org). CSME was defined as CME with less than a 0.2 logMAR CDVA improvement compared with the preoperative baseline.
Small studies have shown that a combination of topical corticosteroids and NSAIDs is more effective than single-drug treatment in patients with acute CSME after cataract surgery.22 Patients who developed CSME during the course of this study were therefore treated with bromfenac 0.09% eyedrops twice daily and dexamethasone 0.1% eyedrops 4 times daily for 4 weeks. If CSME resolved within these 4 weeks, bromfenac was stopped and dexamethasone eyedrops were reduced with 1 drop per day every following week. If CSME persisted after 4 weeks of topical treatment, patients received 1 intravitreal injection of 0.05 mL (1.25 mg) bevacizumab. From 12 weeks postoperatively, CSME was treated in accordance with the standard of care at each participating study center.
Moreover, a stepwise treatment approach was used if patients developed severe postoperative inflammation, defined as at least 2+ cells according to the SUN classification.
The sample-size calculation for this study was based on the primary outcome—the difference in central subfield mean macular thickness at 6 weeks postoperatively. Sample size was recalculated, masked for treatment allocation, based on overall nuisance parameters obtained from the first 159 patients included in the study; that is, a standard deviation (SD) of a postoperative central subfield mean macular thickness of 42.25 μm and a correlation between the mean preoperative and postoperative central subfield macular thickness of 0.62. With a total sample size of 741, a mean difference of 10 μm between treatment groups could be detected with 80% power at a Bonferroni-corrected significance level of 0.0167, where 40 patients were added to correct for center, center with treatment interactions, and other covariates. To compensate for a maximum dropout rate of 20%, the total calculated sample size for this study was 926 patients.
No interim analyses were performed during the course of the study. Data were collected in a web-based database (Macro, version 4.4C) and were subsequently exported to SPSS Statistics for Windows software (version 23.0, IBM Corp.) for analysis. Generalized logistic mixed-model analyses were performed using the GLMER function within RD (version 3.3.1, The R Foundation for Statistical Computing).
Data were excluded from the final analyses if patients developed a perioperative complication that was expected to increase the risk for developing CME.4,23 These included patients with a posterior capsule rupture with or without vitreous loss, zonulysis, or iris trauma. Other perioperative complications, such as intraoperative floppy-iris syndrome without iris trauma, IOL damage, and corneal erosions, were thought not to influence the risk for developing CME. These cases were not excluded from the final analyses.
Categorical data were presented as number of patients (and percentage) and numerical data as means ± SD. Differences between treatment groups in the occurrence of perioperative complications, adverse events, dropout, and reasons for dropout between treatment groups in were assessed using the chi-square test or Fisher exact test, where appropriate.
Postoperative data analyses were performed with masking for the treatment groups. Because of the Bonferroni adjustment for multiple testing, a P value less than 0.0167 was considered statistically significant. Linear mixed-model analyses using restricted maximum likelihood were used for numerical outcomes that were measured repeatedly, including the primary outcome of the study. Linear mixed models account for the correlation between repeated measurements within a patient and also include data from patients who dropped out of the study. No data were imputed. A likelihood-based approach was used for missing outcome data, including SD-OCT data, for which data were assumed to be missing at random. No missing data were expected for the fixed factors included in the model, such as study center, timing of study visits, and the interaction term treatment group × timing of the study visit. Logistic regression analysis was used to check whether missing outcome data were significantly related to certain patient characteristics. Variables significantly related to missing outcome data were added to the model to ensure missing at random. In addition, baseline patient characteristics that are known to be related to the primary outcome, such as sex, age, and the presence of CME in the fellow eye, were included, while the model without these patient characteristics was used as a sensitivity analysis.4,23 A second sensitivity analysis was performed with study center included as a random instead of a fixed factor. A 3-way interaction term treatment group × timing of the study visit × study center (plus their 2-way interaction terms) was applied to verify whether treatment effect depended significantly on study center. If this 3-way interaction term proved significant, the treatment effect was reported for each center separately. If the 3-way interaction term was not significant, it was removed from the model and only the overall treatment effect over all centers was reported.
Generalized logistic mixed-model analyses were used for binary outcomes, in which a random intercept on center level was included to account for the nesting of patients within a center. Adaptive Gauss-Hermite approximation (10 quadrate points) was used for estimating the log-likelihood. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported for pairwise comparisons.
Because the sample size within a center might be small, the above-mentioned analyses were also performed after combining study centers based on location and center type (ie, academic centers in the Netherlands, nonacademic centers in the Netherlands, other West European centers, Central European centers, and South European centers). This was included as a sensitivity analysis.
Finally, per-protocol analyses were performed and compared with the intention-to-treat analyses. Data were excluded from the per-protocol analyses if deviations from the allocated treatment intervention occurred (eg, patients who received additional NSAID or corticosteroid eyedrops).
Figure 2 shows a flow diagram showing the number of participants who were screened, randomized, and analyzed in the study. Baseline characteristics of the participants are shown in Table 1 and Appendix S2 (available at http://jcrsjournal.org).
Twenty-seven patients (3.1%) developed a perioperative complication (Appendix S3, available at http://jcrsjournal.org). There were no significant differences in the number of perioperative complications between treatment groups (P = .668). Data of 8 patients (0.9%) were not included in the final analyses because the reported perioperative complications have been related to an increased risk for developing CME after cataract surgery.
Although the study included only 914 instead of 926 patients, the final dropout rate was substantially lower than the accounted for maximum rate of 20%. Of 906 patients suitable for inclusion in the final data analyses, 86 patients (9.5%) (26 in bromfenac group, 33 in dexamethasone group, and 27 in combination treatment group) were dropouts. There were no statistically significant differences between treatment groups in the number of patients who dropped from the study (P = .560) or reasons for dropout (P = .227). Four patients voluntarily postponed cataract surgery and 40 patients withdrew from study participation before the day of cataract surgery. The 2 patients in the bromfenac group who discontinued participation because of adverse events considered not related to topical NSAID treatment had increased IOP and cerebral metastases of a preexisting lung carcinoma, respectively. Of the 6 patients in the dexamethasone group considered dropouts because of adverse events, 2 patients had red eyes and 1 developed a corneal erosion. One patient in the dexamethasone group was dropped from the study when CSME was diagnosed; however, primary outcome data were available. One patient in the dexamethasone group discontinued participation after developing lymphatic metastases of a preexisting rectal carcinoma, and another patient died after developing a cerebrovascular infarction. No patient in the combination treatment group dropped out because of adverse events.
Macular Thickness and Volume
Patients treated with a combination of bromfenac and dexamethasone eyedrops had the lowest postoperative macular thickness after correction for baseline measurements. (Table 2 and Figure 3). Six weeks postoperatively, the mean central subfield macular thickness was 288.3 μm, 296.0 μm, and 284.5 μm in the bromfenac group, dexamethasone group, and combination treatment group, respectively (overall P = .006). After correction for the mean baseline central subfield macular thickness and other covariables, the mean central subfield macular thickness was 9.6 μm higher in the dexamethasone group than in the combination treatment group (P = .002). It was 6.9 μm higher in the dexamethasone group than in bromfenac group, although the difference between the 2 treatment groups was not statistically significant (P = .026). The mean central subfield mean macular thickness was 2.7 μm higher in the bromfenac group than in the combination treatment group (P = .380). The mean central subfield mean macular thickness was comparable in all treatment groups 12 weeks after cataract surgery (overall P = .299).
Figure 4 shows the parafoveal thickness, perifoveal thickness, total macular volume, and CDVA by group over time. Patients treated with topical dexamethasone had a significantly higher parafoveal thickness at 6 weeks than patients treated with bromfenac (P = .006) or with a combination of bromfenac and dexamethasone (P = .001). The parafoveal thickness was comparable in the bromfenac group and combination treatment group (P = .520). At 12 weeks, the parafoveal thickness was comparable between all treatment groups (overall P = .136).
The postoperative perifoveal thickness was also highest in the dexamethasone treatment group. At 6 weeks, the perifoveal thickness was significantly higher in patients treated with topical dexamethasone than in patients treated with bromfenac or with bromfenac and dexamethasone (P = .012 and P = .008, respectively). There were no significant differences in perifoveal thickness between the bromfenac group and combination group (P = .883). At 12 weeks, the perifoveal thickness was lowest in the bromfenac treatment group. The mean perifoveal thickness was 2.1 μm higher in the dexamethasone group and 1.8 μm higher in the combination treatment group (P = .004 and P = .014, respectively).
Finally, the mean total macular volume was significantly higher in the dexamethasone treatment group than in the other treatment groups 6 weeks postoperatively (P < .001). Up to the end of the study, the total macular volume remained significantly lower in the bromfenac treatment group than in the dexamethasone group (P = .001). After Bonferroni correction, there were no statistically significant differences between the bromfenac group and combination group (P = .045).
Only overall treatment effects are reported because the effects were not significantly different between study centers for any postoperative outcome. Sensitivity analyses showed similar results with respect to total macular volume and macular thickness measurements. Per-protocol analyses showed comparable results (Appendix S4, available at http://jcrsjournal.org).
Incidence of Cystoid Macular Edema and Clinically Significant Macular Edema
Most cases of CME developed within 6 weeks postoperatively (Table 3). There were clinically important and statistically significant differences in the incidence of CME within 6 and 12 weeks postoperatively between treatment groups (overall P = .002 and P = .003, respectively). The odds of developing CME were significantly higher in the dexamethasone treatment group than in the combination group (OR, 4.7; 95% CI, 1.7-12.5; P = .002). The odds were also higher in patients treated with bromfenac eyedrops than in patients in the combination group; however, a pairwise comparison did not reach statistical significance (OR, 2.2; 95% CI, 0.8-6.5; P = .144). A similar pattern was seen 12 weeks postoperatively (Table 3).
Although the incidence of CSME remained highest in the dexamethasone treatment group during the course of the study (Table 3), the differences were statistically significant at 6 weeks only (overall P = .006 and P = .043 within 6 weeks and 12 weeks, respectively).
Sensitivity and per-protocol analyses showed comparable results, and the effects were not significantly different between study centers.
There were no significant differences in CDVA 6 and 12 weeks postoperatively between groups (overall P = .327 and P = .309, respectively) (Table 2). The mean final visual acuity was −0.02 logMAR (21/20 Snellen) in the bromfenac treatment group and −0.01 logMAR (20/20 Snellen) in the other treatment groups.
In patients with CSME, the maximum decline in CDVA was up to 0.54 logMAR (20/70 Snellen). The median CDVA decreased to 0.20 logMAR (20/32 Snellen) but improved to 0.07 logMAR (20/26 Snellen) 12 weeks postoperatively.
All adverse events and serious adverse events are shown in Appendices S5 and S6 (available at http://jcrsjournal.org). No treatment-related serious adverse events were reported during the course of this study. Adverse events occurred in 310 patients. These events were reported in 104 patients (34.3%), 120 patients (39.9%), and 86 patients (28.5%) in the bromfenac group, dexamethasone group, and combination treatment group, respectively (overall P = .013). Frequently reported adverse events were mild pain or foreign-body sensation, dry eyes, tearing, and posterior vitreous detachment.
Six weeks postoperatively, the presence of cells on slitlamp examination was reported in 11 (4.0%) of 272 patients in the bromfenac group, 12 (4.4%) of 271 patients in the dexamethasone group, and in 15 (5.5%) of 273 patients in the combination treatment group. Flare was reported in 1 patient in the combination treatment group. At 12 weeks, anterior chamber cells were reported in 1 (0.4%) of 268 patients in both the bromfenac and combination treatment groups and in 5 (1.9%) of 264 patients in the dexamethasone group. No patient presented with anterior chamber flare at 12 weeks. No patient had 2+ or more cells according to the SUN classification 6 or 12 weeks postoperatively.
The mean IOP decreased from 16.1 ± 3.3 mm Hg at baseline to 13.4 ± 2.9 mm Hg 12 weeks after cataract surgery. There were no statistically significant differences in the mean IOP between treatment groups at 6 or 12 weeks (overall P = .707 and P = .877, respectively). One patient in the bromfenac group developed severely increased IOP and was hospitalized to receive intravenous medication. Four patients in the dexamethasone group and 2 patients in the combination treatment group received topical treatment to reduce IOP; however, none received a prostaglandin analog. Two patients in the dexamethasone group and 1 patient in the combination treatment group received additional oral acetazolamide.
Modern phacoemulsification cataract surgery techniques have substantially decreased the incidence of CME after cataract surgery. However, because phacoemulsification is one of the most frequently performed surgical procedures worldwide, even small improvements in perioperative care can be clinically important. Many studies have evaluated the optimum treatment to prevent CME after cataract surgery; however, most were insufficiently powered to detect small, yet clinically relevant differences between treatment groups. Moreover, most studies did not compare the efficacy of an NSAID only versus combination treatment with a topical corticosteroid. The ESCRS PREMED study included 914 nondiabetic patients and at present is the largest multicenter study directly comparing the efficacy of a topical NSAID, corticosteroid, and combination treatment to prevent CME after cataract surgery.
The overall incidence of CSME in this study was 3.4%, while previous studies have reported a lower incidence of 1.2% to 2.0% in patients with no risk factors for developing CSME after cataract surgery.4,23 However, a fair comparison of incidence rates between studies requires standardization of the applied definitions.9,21 The ESCRS PREMED study definition for CME is very sensitive because of a very low cutoff value with respect to macular thickening; that is, at least a 10% increase in the mean central subfield macular thickness compared with the preoperative baseline. Moreover, SD-OCT allows detection of very small cystoid changes, and some of them might not be clinically relevant. Clinically significant macular edema was defined as the occurrence of CME on OCT and less than 0.2 logMAR improvement in CDVA compared with the preoperative baseline. This definition assumes that modern phacoemulsification cataract surgery leads to at least 2 lines of improvement in CDVA. However, the current study also included patients with a preoperative CDVA of 20/25 Snellen or better who had cataract surgery for decreased quality of vision (eg, glare or decreased contrast sensitivity) because many national guidelines report that a reduced visual acuity is not the only indication for cataract surgery. Therefore, the PREMED study definition for CSME might have caused a surplus of CME cases to be considered clinically significant.
Previous studies have assed the efficacy of topical antiinflammatory treatments to prevent vision loss resulting from CME after cataract surgery.3 The ESCRS PREMED study could not identify a statistically significant difference in the mean postoperative CDVA between treatment groups. Although the CDVA decreased to 20/70 Snellen in patients with CSME, those outliers did not cause significant differences with respect to the mean CDVA 6 weeks and 12 weeks postoperatively. Nevertheless, individual patients will benefit most from optimum prevention of visually significant CME.
The odds ratio of developing CSME within 12 weeks postoperatively was 2.6 in patients treated with topical bromfenac and 3.7 in patients treated with topical dexamethasone compared with the combination treatment group. The incidence of CSME within 12 weeks postoperatively was 3.6%, 5.1%, and 1.5% in the bromfenac group, dexamethasone group, and combination treatment group, respectively. Although these differences were not statistically significant after Bonferroni correction (P = .043), the rates clearly indicate benefit from the use of both bromfenac and dexamethasone eyedrops to prevent CSME after uneventful cataract surgery.
Various topical NSAID preparations are available; however, the U.S. Food and Drug Administration has not approved any for the prevention of CME after cataract surgery.3 The European Medicines Agency (EMA) approved nepafenac for prevention of inflammation and macular edema after cataract surgery in diabetic patients, whereas bromfenac is only approved for prevention of postoperative inflammation. Despite this, many ophthalmologists continue to use bromfenac and other topical NSAIDs to prevent CME after cataract surgery because CME is likely a result of an overall postoperative inflammatory response.3,21 At the initiation of this European multicenter trial in 2012, the ESCRS PREMED Study Group decided to use bromfenac 0.09% because previous studies7,24 indicated better penetration into ocular tissues, extended duration of antiinflammatory activity, and enhanced inhibitory effect on cyclooxygenase 2. The results of an indirect network comparison and several small RCTs suggest some differences between NSAID preparations in their efficacy to reduce inflammation after cataract surgery.25 The quality of evidence, however, is low to moderate, in part because of the small samples in the available studies. Previous studies did not indicate significant differences between various NSAIDs in their efficacy to prevent macular thickening.25,26 More recently, reformulated once-daily NSAID preparations have been shown to be effective in preventing inflammation and CME after cataract surgery.4,27 Therefore, healthcare costs and patient comfort could be taken into account when choosing the optimal NSAID to be used as a standard of care in clinical practice.2,28
In the ESCRS PREMED study, the efficacy of bromfenac eyedrops was compared with that of dexamethasone 0.1%. Dexamethasone was the corticosteroid of choice because this preparation was used at 8 of 9 PREMED study centers that included a topical corticosteroid in their standard of care. A previous metaanalysis10 showed that both potent steroids and weaker steroids are less effective than topical NSAIDs in preventing CME after cataract surgery.
The primary outcome of the ESCRS PREMED study was the difference in the mean postoperative central subfield macular thickness. The results suggest that topical bromfenac 0.09% is more effective than topical dexamethasone 0.1% in preventing retinal thickening after cataract surgery. These results are in line with the conclusions of previous systematic reviews.9–11 The PREMED study found no statistically significant or clinically relevant differences in the mean macular thickness and total macular volume 6 weeks postoperatively between the bromfenac group and combination group. The odds ratio for developing CSME within 12 weeks postoperatively was 2.6 in patients treated with topical bromfenac when compared to patients receiving combination treatment; however, the differences were not statistically significant. In this study, topical bromfenac was used twice daily for 2 weeks postoperatively, as per the EMA Summary of Product Characteristics 2011.29 Although the EMA recommends topical bromfenac treatment for 2 weeks postoperatively, ophthalmologists have studied its use for up to 2 months after cataract surgery. One small study of 72 patients30 compared the efficacy of topical bromfenac for 2 months postoperatively versus a combination treatment with a corticosteroid. The study found no significant differences in postoperative inflammatory response and reported no cases of CME in the bromfenac group or the combination treatment group. However, the sample size in this study might have been too small and the study might have had insufficient power to detect clinically relevant differences. Another study31 reported no added value of using bromfenac eyedrops for more than 1 month postoperatively. Nevertheless, future RCTs are needed to compare the efficacy of topical bromfenac for 1 month postoperatively versus combination treatment with a corticosteroid to prevent the occurrence of CME and CSME after cataract surgery.
This European RCT has limitations that are related to its study design and the multinational setting. First, macular thickness and total macular volume were measured using 5 different SD-OCT devices. Where some OCT devices measure retinal thickness from the internal limiting membrane to the retinal pigment epithelium (RPE), others use Bruch membrane as the outer retinal boundary. These differences are most relevant in pathologies that affect the RPE–Bruch membrane complex.32 Because CME is caused by an accumulation of fluid in the inner nuclear and outer plexiform layers of the retina, all OCT devices will detect postoperative changes in macular thickness and total macular volume. Each study center used only 1 type of SD-OCT device for all preoperative and postoperative measurements. Furthermore, differences between treatment groups reported in this study were corrected for baseline measurements. Second, the current study had a follow-up of 3 months postoperatively. The currently available literature does not provide clear evidence with regard to the long-term efficacy of topical NSAIDs and/or corticosteroids on CDVA more than 3 months after regular cataract surgery, although some studies provide long-term results in high-risk patients (eg, patients with diabetes mellitus or uveitis).11 The current study has shown a benefit to using a combination of topical bromfenac and dexamethasone eyedrops up to 3 months postoperatively. Considering that most cases of CME develop within 4 to 6 weeks after cataract surgery, preventive treatments should be optimized in the immediate postoperative period. However, further studies are needed to determine whether a combination of topical bromfenac and dexamethasone can also prevent the occurrence of CME and CSME after 3 months postoperatively. Finally, this study was designed to evaluate the optimum preventive strategy for CME in nondiabetic patients with no other risk factors for developing CME. Therefore, the results in this study might not be applicable to patients with diabetes, uveitis, retinal vein occlusion, epiretinal membrane, or other risk factors for developing CME and in cases of complicated cataract surgery.4,23
Future work will evaluate the cost-effectiveness of topical NSAIDs, corticosteroids, and combination treatments after cataract surgery. When choosing the optimum NSAID eyedrop for standard care, patient satisfaction, simplicity of drug administration, frequency of drug administration, and ocular comfort should be taken into account. Although the efficacy of various topical NSAIDs seems to be comparable, their cost-effectiveness should be studied in more detail.
In conclusion, the ESCRS PREMED study evaluated the optimum treatment regimen to prevent the occurrence of CME after uneventful phacoemulsification cataract surgery in nondiabetic patients with no other risk factors for developing CME. According to the results in this randomized controlled European multicenter trial, the postoperative macular thickness and total macular volume were lower in patients treated with topical bromfenac 0.09% than in patients treated with topical dexamethasone 0.1%. A combination treatment of dexamethasone and bromfenac eyedrops could not further reduce postoperative macular thickness, total macular volume, or CDVA compared with topical bromfenac alone. However, the odds of developing CSME within 12 weeks postoperatively were lowest in the combination treatment group.
What Was Known
- The decision on standard postoperative care could have significant effects on healthcare costs.
- Small studies have shown that as compared with topical corticosteroids alone, a combination of topical NSAIDs and corticosteroids significantly reduces the odds of developing CME after cataract surgery.
What This Paper Adds
- A combination of topical bromfenac 0.09% and dexamethasone 0.1% was not significantly more effective than topical bromfenac alone in reducing the mean postoperative macular thickness and total macular volume; however, the overall incidence of CME was lower in the combination treatment group.
- The odds ratio for developing clinically significant CME within 12 weeks postoperatively were 2.6 in patients treated with topical bromfenac, and 3.7 in patients treated with topical dexamethasone when compared to patients treated with a combination of both drugs.
1. Bourne RRA, Stevens GA, White RA, Smith JL, Flaxman SR, Price H, Jonas JB, Keeffe J, Leasher J, Naidoo K, Pesudovs K, Resnikoff S, Taylor HR., on behalf of the Vision Loss Expert Group. (2013). Causes of vision loss worldwide, 1990–2010: a systematic analysis. Lancet Glob Health, 1
, e339-e349, Available at: http://www.thelancet.com/pdfs/journals/langlo/PIIS2214-109X(13)70113-X.pdf
2. World Health Organization. Vision 2020. The Right to Sight. Global Initiative for the Elimination of Avoidable Blindness; Action Plan 2006–2011. 2007, WHO, Geneva, Switzerland, Available at: http://apps.who.int/iris/bitstream/10665/43754/1/9789241595889_eng.pdf
3. Kim SJ, Schoenberger SD, Thorne JE, Ehlers JP, Yeh S, Bakri SJ. (2015). Topical nonsteroidal anti-inflammatory drugs and cataract surgery; a report by the American Academy of Ophthalmology (Ophthalmic Technology Assessment). Ophthalmology, 122
, 2159-2168, Available at: http://www.aaojournal.org/article/S0161-6420(15)00464-9/pdf
4. McCafferty S, Harris A, Kew C, Kassm T, Lane L, Levine J, Raven M. Pseudophakic cystoid macular edema prevention and risk factors; prospective study with adjunctive once daily topical nepafenac 0.3% versus placebo. BMC Ophthalmol. 17, 2017, 16, Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5319126/pdf/12886_2017_Article_405.pdf
5. Irvine SR. A newly defined vitreous syndrome following cataract surgery; interpreted according to recent concepts of the structure of the vitreous; the Seventh Francis I. Proctor Lecture. Am J Ophthalmol
6. Gass JDM, Norton EWD. Cystoid macular edema and papilledema following cataract extraction; a fluorescein funduscopic and angiographic study. Arch Ophthalmol
7. Hoffman RS, Braga-Mele R, Donaldson K, Emerick G, Henderson B, Kahook M, Mamalis N, Miller KM, Realini T, Shorstein NH, Stiverson RK, Wirostko B., for the ASCRS Cataract Clinical Committee and the American Glaucoma Society. (2016). Cataract surgery and nonsteroidal antiinflammatory drugs. J Cataract Refract Surg, 42
, 1368-1379, Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5531861/pdf/nihms878731.pdf
8. Kim SJ, Patel SN, Sternberg P Jr. (2016). Routine use of nonsteroidal anti-inflammatory drugs with corticosteroids in cataract surgery: beneficial or redundant? [editorial] Ophthalmology, 123
, 444-446, Available at: http://www.aaojournal.org/article/S0161-6420(15)01264-6/pdf
9. Wielders LHP, Lambermont VA, Schouten JSAG, van den Biggelaar FJHM, Worthy G, Simons RWP, Winkens B, Nuits RMMA. Prevention of cystoid macular edema after cataract surgery in nondiabetic and diabetic patients: a systematic review and meta-analysis. Am J Ophthalmol
10. Kessel L, Tendal B, Jφrgensen KJ, Erngaard D, Flesner P, Lundgaard Andresen J, Hjortdal J. (2014). Post-cataract prevention of inflammation and macula edema by steroid and nonsteroidal anti-inflammatory eye drops; a systematic review. Ophthalmology, 121
, 1915-1924, Available at: http://www.aaojournal.org/article/S0161-6420(14)00389-3/pdf
11. Lim BX, Lim CHL, Lim DK, Evans JR, Bunce C, Wormald R. Prophylactic non-steroidal anti-inflammatory drugs for the prevention of macular oedema after cataract surgery. Cochrane Database Syst Rev. issue 11, 2016, Article no. CD006683. Available at: http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD006683.pub3/epdf
12. Juthani VV, Clearfield E, Chuck RS. Non-steroidal anti-inflammatory drugs versus corticosteroids for controlling inflammation after uncomplicated cataract surgery. Cochrane Database Syst Rev. issue 7, 2017, article no. CD010516. Summary available at: http://onlinelibrary.wiley.com/doi/10.1002/14651858.CD010516.pub2/epdf/abstract
13. World Health Organization. International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10). 2016, WHO, Geneva, Switzerland, Available at: http://apps.who.int/classifications/icd10/browse/2016/en
14. Donnenfeld ED, Perry HD, Wittpenn JR, Solomon R, Nattis A, Chou T. Preoperative ketorolac tromethamine 0.4% in phacoemulsification outcomes: pharmacokinetic-response curve. J Cataract Refract Surg
15. Yavas GF, Öztürk F, Küsbeci T. Preoperative topical indomethacin to prevent pseudophakic cystoid macular edema. J Cataract Refract Surg
16. Chylack LT Jr, Leske MC, McCarthy D, Khu P, Kashiwagi T, Sperduto R. Lens Opacities Classification System II (LOCS II). Arch Ophthalmol
17. The Standardization of Uveitis Nomenclature (SUN) Working Group. Standardization of uveitis nomenclature for reporting clinical data. Results of the First International Workshop. Am J Ophthalmol
18. Holladay JT, Prager TC. Mean visual acuity [letter]. Am J Ophthalmol
19. Early Treatment Diabetic Retinopathy Study Research Group. Early Treatment Diabetic Retinopathy Study design and baseline patient characteristics; ETDRS report number 7. Ophthalmology
20. Degenring RF, Vey S, Kamppeter B, Budde WM, Jonas JB, Sauder G. Effect of uncomplicated phacoemulsification on the central retina in diabetic and non-diabetic subjects. Graefes Arch Clin Exp Ophthalmol
21. Yonekawa Y, Kim IK. Pseudophakic cystoid macular edema. Curr Opin Ophthalmol
22. Wielders LHP, Schouten JSAG, Aberle MR, Lambermont VA, van den Biggelaar FJHM, Winkens B, Simons RWP, Nuijts RMMA. Treatment of cystoid macular edema after cataract surgery. J Cataract Refract Surg
23. Chu CJ, Johnston RL, Buscombe C, Sallam AB, Mohamed Q, Yang YC., for the United Kingdom Pseudophakic Macular Edema Study Group. (2016). Risk factors and incidence of macular edema after cataract surgery; a database study of 81 984 eyes. Ophthalmology, 123
, 316-323, Available at: http://www.aaojournal.org/article/S0161-6420(15)01146-X/pdf
24. Carreño E, Portero A, Galarreta DJ, Herreras JM. (2012). Update on twice-daily bromfenac sodium sesquihydrate to treat postoperative ocular inflammation following cataract extraction. Clin Ophthalmol, 6
, 637-644, Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3346189/pdf/opth-6-637.pdf
25. Duan P, Liu Y, Li J. The comparative efficacy and safety of topical non-steroidal anti-inflammatory drugs for the treatment of anterior chamber inflammation after cataract surgery: a systematic review and network meta-analysis. Graefes Arch Clin Exp Ophthalmol
26. Lee TH, Choi W, Ji YS, Yoon KC. (2016). Comparison of ketorolac 0.45% versus diclofenac 0.1% for macular thickness and volume after uncomplicated cataract surgery. Acta Ophthalmol, 94
, e177-e182, Available at: http://onlinelibrary.wiley.com/doi/10.1111/aos.12780/pdf
27. Walters TR, Goldberg DF, Peace JH, Gow JA., for the Bromfenac Ophthalmic Solution 0.07% Once Daily Study Group. (2014). Bromfenac ophthalmic solution 0.07% dosed once daily for cataract surgery; results of 2 randomized controlled trials. Ophthalmology, 121
, 25-33, Available at: http://www.aaojournal.org/article/S0161-6420(13)00595-2/pdf
28. Wilson DJ, Schutte SM, Abel SR. Comparing the efficacy of ophthalmic NSAIDs in common indications: a literature review to support cost-effective prescribing. Ann Pharmacother
29. European Medicines Agency. Yellox, bromfenac. First published 14/06/2011; updated 17/11/2017. Available at: http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/001198/human_med_001455.jsp&mid=WC0b01ac058001d124
30. Miyanaga M, Miyai T, Nejima R, Maruyama Y, Miyata K, Kato S. (2009). Effect of bromfenac ophthalmic solution on ocular inflammation following cataract surgery. Acta Ophthalmol, 87
, 300-305, Available at: http://onlinelibrary.wiley.com/doi/10.1111/j.1755-3768.2008.01433.x/pdf
31. Wang Q-w, Yao K, Xu W, Chen P-q, Shentu X-c, Xie X, Weng Y, Zhang L, Jin C-f, Wu W, Zhu Y-n, Yu Y-h. Bromfenac sodium 0.1%, fluorometholone 0.1% and dexamethasone 0.1% for control of ocular inflammation and prevention of cystoid macular edema after phacoemulsification. Ophthalmologica
32. Giani A, Cigada M, Choudhry N, Deiro AP, Oldani M, Pellegrini M, Invernizzi A, Duca P, Miller JW, Staurenghi G. Reproducibility of retinal thickness measurements on normal and pathologic eyes by different optical coherence tomography instruments. Am J Ophthalmol
None of the authors has a financial or proprietary interest in any material or method mentioned.
Supplemental material available at www.jcrsjournal.org.
Other cited material
A. U.S. National Institutes of Health Clinical Trials. PRevention of Macular EDema After Cataract Surgery (PREMED). NCT01774474. Available at: https://clinicaltrials.gov/ct2/show/NCT01774474
. Accessed February 17, 2018
B. ALEA, version 3.0, Formsvision BV, Abcoude, The Netherlands. Available at: http://www.aleaclinical.eu/
. Accessed February 17, 2018
C. MACRO, Advanced data collection for clinical research. Amsterdam, the Netherlands, Elsevier, 2018. Available at: https://www.elsevier.com/solutions/macro
. Accessed February 17, 2018
D. GLMER. The R Foundation for Statistical Computing. Available at: https://www.rdocumentation.org/packages/twidlr/versions/0.0.0.9000/topics/glmer
. Accessed February 17, 2018