A prospective, randomized, double-masked, placebo-controlled clinical trial was performed at two sites in Tucson Arizona, as part of the Arizona Surgical Eye Study (ASES). The ASES was an independent prospective randomized study designed to examine the incidence and causation of post cataract surgical complications conducted from 10–15-2013 to 11–1–2015 stemming from NSAID use, phacoemulsification machine use, IOL type, and identified risk factors. The ASES specifically examined complication type and incidence, ocular discomfort, inflammation, capsular phimosis, endothelial cell count change, and posterior capsule opacification postoperatively. The clinical evaluation described herein is a subset of the available data at the time of publication examining solely the incidence of PCME. The clinical trial was approved by Western Independent Review Board. All patients were treated according to the Declaration of Helsinki document on human research ethics, and underwent informed consent.
Study protocol
Subjects, 18 and older, were enrolled from the clinic. The planned enrollment was 970 eyes (1000 completed) for probable statistical significance with a 1:1 ratio of control: treatment. Subjects were chosen from patients who had visually significant cataracts and were to undergo phacoemulsification with implantation of an intracapsular positioned intraocular lens. Subject’s eyes individually were randomly assigned by the compounding pharmacy using random number generator in groups of 10 to receive a placebo of sterile saline drops or nepafenac 0.3%. Both study drops (nepafenac and placebo) were produced individually by the pharmacy in a generic bottle marked by a code and instructions for use. The codes were maintained and utilized solely by the pharmacy to determine the content of the bottle revealed following the completion of the study. All patients received topical prednisolone 1% four times daily for the first week, tapered to 2 times daily over the second week and 1 time daily for the subsequent 3 weeks which approximates most common practice [
1,
3,
4]. No additional steroids were used intraoperatively. Ofloxacin 0.3% was used 4 times daily for the first week and discontinued.
In contrast to other studies, patients included those on prostaglandins or on glaucoma medications, those with an epiretinal membrane, macular degeneration, or diabetes mellitus (with or without retinopathy), macular hole, previous retinal surgery, and history of central or branch retinal vein occlusion. Glaucoma medication patients were included as a risk factor if the medications were used any time prior to the surgery. Epiretinal membranes were identified and included if seen on funduscopic exam or OCT prior to cataract surgery within the macula (central 5.5 mm diameter around the fovea or 2.75 disk diameters). Diabetic retinopathy was included if there was any demonstrated evidence upon funduscopic exam. Previous retinal surgery included any patient receiving any vitreoretinal procedure excluding laser procedures. Macular holes, central and branch retinal vein occlusions were included regardless of stage. Exclusion criteria included previous uveitis (<1 year), previous anterior segment intraocular surgery or a hypersensitivity or allergy to NSAIDs. Any patients who had a complicated cataract surgery (posterior capsule rupture, vitreous loss, retained cortical material, significant corneal edema or an IOL not placed in the capsular bag) were excluded from the data analysis.
During the patient’s initial visit, information on demographics was obtained, including age, sex, birth date, ethnicity, as well as ocular history and medical history. The baseline exam included intraocular pressure, dilated fundus exam, slitlamp exam and best-corrected visual acuity (BCVA) by Snellen chart. Also, before surgery, a trained technician performed a baseline OCT macular cube scan (OCT protocol described below).
Three surgeons participated, LL, SM, and JL, with similar methods and amount of time performing the surgery. The surgeries were phacoemulsification cataract extraction with posterior chamber in-the-bag IOL placement. An equal number of patients (500 in each group) were randomized in groups of 10 and underwent phacoemulsification cataract surgery with an Alcon (Fort Worth, TX) Infinity machine using an SA60AT IOL and an Abbott Medical Optics (Santa Ana, CA) WhiteStar Signature using a ZCB00 IOL. Both the NSAID treatment group and control group were equally divided among the two phacoemulsification machines and their respective IOL’s which was confirmed following the study. Postoperatively, patients were routinely followed up on day one, at 1 week and at 5 weeks. All visits included a BCVA, slit lamp exam, IOP, and an OCT macular cube scan. Subjects were assessed at all visits for any postoperative inflammation, capsular phimosis and posterior capsule opacification. During the study, patients were interviewed about any issues or concerns they had during that postoperative period including medicine cost [
19]. Any adverse events or complications were followed closely and treated during the post-operative period.
Once the two groups were randomized, one group received one drop a day of nepafenac 0.3% for 5 weeks which is the recommended dosage for the new higher concentration nepafenac. Nepafenac ophthalmic suspension 0.3% was supplied as a sterile, aqueous suspension with a pH approximately of 6.8. The osmolality of the nepafenac 0.3% is approximately 300 mOsm/kg. Nepafenac 0.3% contains: Active: nepafenac 0.3% Inactive: boric acid, propylene glycol, carbomer 974P, sodium chloride, guar gum, carboxymethylcellulose sodium, edentate disodium, benzalkonium chloride 0.005%. The topical NSAID was begun on the first post-operative day of surgery. The other group received the placebo which were buffered sterile saline drops with a pH approximately of 7.0 and osmolality of 290 mOsm/kg, dosed at the same frequency. The placebo drops ingredients included: Sodium chloride; Boric acid; calcium chloride; magnesium chloride; potassium chloride; purified water; sodium borate; and carboxymethylcellulose sodium. The pharmacy provided identical opaque generic bottles that were coded with identification numbers only; no drug information was visible on the bottle. The bottles also had a patient identification number and expiration date displayed. Only the pharmacy retained the information for the codes and this information was not shared with investigators until completion of the study. All participants of the study also received prednisolone acetate 1% suspension 4 times a day for 1 week, followed by 2 times a day for 1 week, followed by 1 time daily for 3 weeks. These medications were started on the day after the surgery. Ofloxacin 0.3% drops were administered to all participants and dosed at four times a day, starting one day before surgery until 1 week after surgery.
Study endpoint
Post-operative clinical findings of PCME within 6 weeks including symptomatic or decreased BCVA and angiographic as well as OCT corroboration to confirm the presence of PCME with experienced vitreo-retinal surgeons. Clinically significant PCME was defined as including both: 1. A loss of two (2) lines of best corrected visual acuity (BCVA) from the expected post cataract BCVA (example: 20/40 BCVA from an expected 20/20 BCVA) or visually symptomatic distortion. 2. OCT, clinical, and angiographic demonstrated PCME. All cases of PCME were successfully treated by the retinal surgeon with continuation of drops and subtenons injections of steroids. Secondary outcome measurements included ocular discomfort, inflammation as well as the comparison of ophthalmic medications for tolerability (COMTOL) questionnaire analysis. The intent-to-treat group consisted of subjects who completed surgery, received the study drugs and completed the follow-up. However, any and all patients who received the study drugs were included in the safety analysis. Complications and adverse events were documented when interviewing the participants or recorded by investigators. Adverse events, for the purpose of this study, were defined as any unintended or unwanted sign, symptom or clinical result that could be associated with the study drugs.
Statistical analysis
The sample size using the Wald method was determined prior to the study at approximately 485 in each group. Given the likely incidence of clinically significant PCME between 1 and 3% this sample size would likely lead to a statistically significant outcome [
18]. If no statistical difference was seen at this sample size, then it would be expected that a routine cataract surgeon would not see any difference in his patient outcomes in over 2 years of surgery (@500 cases/year) [Arizona Surgical Eye Study: Survey of 62 U.S. cataract surgeons regarding their use of topical NSAID’s following cataract surgery May to July 2014, unpublished]. A 35% increase in macular volume is considered to be significant by a previous study and was used as our threshold for the OCT measures [
10]. Comparison of OCT’s between five (5) weeks and baseline macular volumes were performed within subjects using a 2-way ANOVA (
p = .05) between subjects in placebo and treatments groups. Two additional 2-way ANOVAs were also completed within subjects pre-op and 5-week OCTs between placebo and treatment groups, both with and without risk factors. As mentioned previously, a loss of more than 2 lines of expected BCVA with angiographic evidence of CME was used to define “clinically significant” PCME. The incidence of clinically significant PCME with both NSAID nepafenac 0.3% treatment and control was determined and compared with Chi-squared (Dof = 1) tests with all patients and both including and excluding risk factors. An odds ratio or relative risk of probable PCME was calculated for each of the identified risk factors by comparing the ratio of the incidence of PCME found with each risk factor to the baseline incidence of PCME without risk factors. Lastly, a logistical regression analysis was completed to examine the probability of multiple other factors (such as IOL type, phaco machine used) affecting the outcome of the study.