The Steroids for Corneal Ulcers Trial:  Study Design and Baseline Characteristics FREE

Infectious corneal ulcers are a major cause of vision loss, with an annual occurrence estimated conservatively at 1.5 to 2 million globally.¹ The World Health Organization estimates that corneal opacities, including corneal ulceration, are the fourth leading cause of blindness.² Studies in South India indicate that approximately half of corneal ulcers are bacterial, and this proportion is typically higher in the United States and Europe.^3- 6 Clearance of the infectious agent in bacterial keratitis is usually successful, yet visual outcomes may be poor. Scarring that accompanies the resolution of infection is thought to be a major contributor to visual impairment.

It has been debated whether topical corticosteroids along with antibiotics reduce immune-mediated tissue damage and improve clinical outcomes. Potential negatives of corticosteroids include worsening of the infection, corneal thinning, perforation, increased intraocular pressure, and cataract development. A lack of evidence exists on the effect of topical corticosteroids for bacterial keratitis. There have been 3 small clinical trials that studied the effect of topical corticosteroids, but these were not powered to achieve statistical significance.^7- 9

To provide evidence on whether the use of adjunctive topical corticosteroids results in better visual outcomes, we undertook a randomized, double-masked, placebo-controlled trial comparing outcomes in patients receiving topical corticosteroids or placebo in addition to topical antibiotics. This report describes the design of the trial and also provides demographic and clinical characteristics of patients with bacterial corneal ulcers from South India and the United States.

The Steroids for Corneal Ulcers Trial (SCUT) is a National Eye Institute–supported, randomized, double-masked, placebo-controlled, comparative, multicenter clinical trial with 2 arms, comparing clinical outcomes in patients with bacterial corneal ulcers receiving topical moxifloxacin, 0.5% (Vigamox; Alcon, Fort Worth, Texas) and topical prednisolone phosphate, 1% (Bausch & Lomb Pharmaceuticals, Inc, Tampa, Florida) or topical moxifloxacin, 0.5%, and topical placebo (sodium chloride, 0.9%, and preservative, prepared by Leiter's Pharmacy, San Jose, California). Institutional review board approval was granted by the Aravind Eye Hospitals institutional review board, the Dartmouth-Hitchcock Medical Center Committee for Protection of Human Subjects, and the University of California, San Francisco, Committee on Human Research. Informed consent was obtained from all subjects.

The primary aim of the study is to determine whether the addition of topical corticosteroids to the treatment of bacterial corneal ulcers improves best spectacle-corrected visual acuity (BSCVA) at 3 months after enrollment. Secondary aims include determining if adjunctive topical corticosteroids affect the following outcomes: adverse events including corneal perforation; size of the infiltrate/scar and rigid contact lens–corrected visual acuity at 3 weeks, 3 months, and 12 months after enrollment; time to resolution of the epithelial defect; and BSCVA at 3 weeks and 12 months. An additional aim is to assess the correlation between minimum inhibitory concentration to moxifloxacin and clinical outcomes. Subgroup analyses will also be performed to assess outcomes across groups based on causative organism, duration of symptoms prior to enrollment, baseline visual acuity, baseline infiltrate/scar size, and location and depth of ulcer.

A sample size of 500 patients (250 per arm) was estimated to have 80% power to detect a 0.20-logMAR (2 lines of visual acuity) difference in BSCVA 3 months after enrollment between the 2 study arms, with a 2-tailed α of .05 and assuming a 20% loss to follow-up. The sample size calculation was based on data from the SCUT pilot study,⁸ with an SD of 0.65 logMAR and controlling for the correlation between 3-month and enrollment BSCVA.

Eligible patients had a culture-positive bacterial corneal ulcer with no evidence of fungal, acanthamoebal, or herpetic keratitis (Table 1). Participants were randomized in a 1:1 ratio either to placebo drops or prednisolone phosphate drops using permuted blocks within study centers. Block sizes were randomized in sizes of 4, 6, and 8. Double-masking was achieved because the prednisolone phosphate solution was identical to placebo. Only the study biostatisticians were unmasked. Enrollment centers included the Aravind Eye Care System (Madurai, Tirunelveli, and Coimbatore, India), Dartmouth-Hitchcock Medical Center, and the Francis I. Proctor Foundation at the University of California, San Francisco.

Patients were randomized to corticosteroid or placebo after they had a confirmed bacterial culture from the cornea and received 48 hours of topical moxifloxacin. The moxifloxacin treatment regimen consisted of 1 drop applied every hour while awake for the first 48 hours, then 1 drop applied every 2 hours until reepithelialization, and then 4 times a day until 3 weeks from enrollment. The corticosteroid regimen consisted of 1 drop applied topically 4 times per day for 1 week after randomization, then twice a day for 1 week, and then once per day for 1 week. Placebo drops were given according to the same schedule. Treating physicians were allowed to stop or change medications at any point during the treatment of the ulcer if they felt it was medically necessary. Study medications were prepared in the United States and shipped to India.

Assessments of BSCVA and infiltrate/scar size were performed at enrollment, 3 weeks, 3 months, and 12 months. Epithelial defect size was measured every 3 days ±1 day from presentation until reepithelialization. In addition to BSCVA, rigid contact lens–corrected visual acuity was measured at 3 weeks, 3 months, and 12 months. Visual acuity was measured by refractionists certified for the study, using a protocol adapted from the Age-Related Eye Disease Study using a tumbling “E” chart at 4 m and logMAR visual acuity (charts 2305 and 2305A; Precision Vision, La Salle, Illinois) (eAppendix).¹⁰ If a patient read fewer than 10 letters at 4 m, acuity was measured at 1 m. If fewer than 10 letters were read at 1 m, low vision was assessed by counting fingers, hand motions, light perception, and no light perception.

A calibrated slitlamp biomicroscope (900; Haag-Streit, Koeniz, Switzerland) was used to assess the size of the infiltrate/scar, epithelial defect, depth, hypopyon, and ocular adverse events including corneal perforation. Infiltrate/scar size and epithelial defect were assessed by measuring the longest dimension and the longest perpendicular to the first measurement, a protocol adapted from the Herpetic Eye Disease Study.¹¹ These measurements were read to the nearest 0.1 mm. As in the Herpetic Eye Disease Study, no differentiation was made between infiltrate and scar. Reepithelialization was defined as the absence of an epithelial defect with administration of fluorescein. Depth was measured in thirds (>0%-33%, >33%-67%, or >67%-100%). All study ophthalmologists were certified for this study.

Photographs were used to determine the location and size of ulcers using newly developed cornea-specific software that superimposed an artificial 4-mm pupil on the photograph, facilitating the grading of the location. A Nikon (Tokyo, Japan) D-series digital SLR camera with a 105-mm f/2.8D AF Micro Nikkor Autofocus Lens and a modified Nikon SB29s electronic flash or Nikon R1 Wireless Close-up Speedlight system was used for corneal photography, performed at enrollment, 3 weeks, 3 months, and 12 months.

Corneal scraping was performed after slitlamp examination at presentation. Two scrapings were smeared for gram stain and potassium hydroxide wet mount. Three scrapings were inoculated onto sheeps' blood agar, chocolate agar, and potato dextrose agar or Sabouraud agar. The criterion for a positive bacterial culture was growth of the organism on 1 solid medium at the site of inoculation. For Staphylococcus epidermis and diphtheroids, cultures were considered positive only if moderate growth was seen on at least 2 solid media or on 1 solid medium plus a gram-stained corneal smear.¹² All patients were checked for fungal elements on smear and culture. Any evidence of fungal infection resulted in exclusion. Repeated cultures were not part of the study protocol but were allowed if deemed necessary by the treating ophthalmologist. Antibiotic susceptibility testing was performed using the Etest method (AB BIODISK, Solna, Sweden), as was done in the SCUT pilot study.¹³ Quality control was performed according to the National Committee for Clinical Laboratory Standards performance standards, recommendations, guidelines, and reports.¹⁴ All microbiologists were certified for the study.

The primary analysis was prespecified as BSCVA at 3 months, analyzed using a linear regression model controlling for enrollment BSVCA. The primary analysis is intention to treat and only includes visits that fall within the 3-month visit window (2.5-5 months from enrollment). Three-week and 12-month BSCVA, rigid contact lens–corrected visual acuity, and infiltrate/scar size will be analyzed in a similar fashion. Infiltrate/scar size and epithelial defect were calculated as the geometric mean of the 2 principal axes in millimeters. Time to reepithelialization is defined as the midpoint between the last observed date with an epithelial defect and the first visit with no epithelial defect and will be analyzed using a Cox proportional hazards model, adjusting for baseline epithelial defect size. Visual acuity in logMAR was used for statistical analyses. Acuities worse than 1.6 logMAR (about 20/800) were recorded as counting fingers (1.7 logMAR), hand motions (1.8 logMAR), light perception (1.9 logMAR), and no light perception (2.0 logMAR), as in the Herpetic Eye Disease Study.¹¹ Corneal perforations between the 2 groups will be compared using a Fisher exact test. For visual acuity measurements taken after therapeutic penetrating keratoplasty, we used the last observation carried forward or 1.7 logMAR acuity, whichever was worse. For infiltrate/scar size following therapeutic penetrating keratoplasty, last observation carried forward was used.

In this report, baseline data are summarized for the study population as a whole. Continuous variables were analyzed with a Wilcoxon rank sum test. Categorical variables were analyzed with a Fisher exact test. All analyses were performed using STATA version 10.0 (StataCorp, College Station, Texas).

Five hundred patients were enrolled between September 11, 2006, and February 22, 2010. Of these patients, 485 (97%) were enrolled in India (Table 2). The most common reason for ineligibility was impending perforation (316 of 1259, 25%) (Table 3). The majority of patients enrolled in the trial were agricultural manual laborers (220 of 500, 44%). The most common object of injury was vegetative matter. Eight patients were contact lens wearers (Table 4).

Median visual acuity was 0.84 logMAR (approximate Snellen equivalent, 20/125) (interquartile range, 0.36-1.7 logMAR; approximate Snellen equivalent, 20/50 to counting fingers) (Table 4). Baseline visual acuity was not significantly different between the United States and India (P = .55). Median infiltrate/scar size was 2.7 mm (interquartile range, 1.9-4.1 mm). Infiltrate/scar size in India was significantly larger than in the United States (P = .04). Ulcers in India were graded as significantly deeper than those in the United States (P = .04) and more likely to be in a central location (P = .002). Coexisting conditions included dacryostenosis/dacryocystitis (103 of 500, 21%) and ocular surface disease (43 of 500, 9%). Patients with corneal ulcer at Aravind Eye Hospital are checked routinely for tear duct obstruction. Six patients had 2 distinct bacterial isolates on culture, so a total of 506 bacterial isolates were obtained in this study (Table 5). The most common organisms isolated were Streptococcus pneumoniae (250 of 506; 49%), Pseudomonas aeruginosa (111 of 506; 22%), and Nocardia species (56 of 506, 11%). The most common organism in the United States was P aeruginosa (5 of 16, 31%).

Use of corticosteroids in the treatment of bacterial corneal ulcers remains controversial, with no definitive evidence to guide treatment decisions.^{7- 8,15- 16} Corticosteroids, if applied in conjunction with appropriate antibacterial therapy, may reduce inflammation that is thought to result in ocular damage including scarring of the cornea.^17- 18 While animal studies show corticosteroids do not directly inhibit the activity of antibiotics, they have been shown to increase the severity of keratitis when administered in the absence of antibiotic therapy.^19- 21 To our knowledge, the SCUT is the first large, prospective randomized clinical trial assessing the impact of adjunctive topical corticosteroids in patients with bacterial corneal ulcers.

Visual acuity was selected as the primary outcome in this trial because it is the most important, clinically relevant long-term outcome for patients. Previous corneal ulcer studies have focused on time to reepithelialization, treatment failure, or poor outcome (eg, increase or no change in infiltrate/scar size, poor final visual acuity, perforation, and/or requiring surgical intervention).^7,22- 24 Corticosteroids may increase the healing time of the ulcer⁸ but still have a beneficial effect on vision. While other outcomes such as scar size or density may be affected by corticosteroids, we believe that visual acuity is the most objective assessment and will take into account these factors.

Corneal ulcers disproportionately affect developing countries; the annual incidence rate has been reported at 113 per 100 000 person-years in Madurai, South India, compared with 27 per 100 000 person-years in Northern California.^25- 26 This makes a sufficiently powered trial much more feasible with the majority of patients enrolled in India. However, this may have implications for generalizability. Contact lens wear is a common risk factor for ulcers in the United States, in contrast to agricultural work in India. The distribution of organisms was statistically different between the United States and India, but all 5 types of bacteria isolated from US ulcers could be found in the top 8 most common bacterial isolates from India.

The ulcers enrolled in this study in India may have been more severe on average than ulcers in the United States. Median baseline visual acuity in the United States was better than in India; however, this difference was not significant. There was a wide range of baseline visual acuities. More than a quarter of patients were enrolled with low visual acuity (counting fingers or worse). More than 20% of patients had visual acuity of 20/40 or better at enrollment, indicating that less severe ulcers were also represented in the trial. Ulcers in the United States had a significantly smaller baseline infiltrate/scar size than those in India. Corneal ulcers from India were more likely to be central in location and to have a deeper infiltrate compared with ulcers from the United States. This variability will facilitate subgroup analyses by ulcer severity, and it is expected that randomization will result in balanced baseline characteristics between the treatment groups.

This report provides comprehensive methods for a corneal ulcer clinical trial. The strengths of this trial include the large sample size and a randomized, masked, controlled trial method; all patients were prospectively followed up with standardized treatment and standardized assessments at prespecified times. The results of this trial will guide treatment practices regarding corticosteroid use in bacterial corneal ulcers. In addition, the data set and specimen bank generated from conducting this trial will enable a multitude of additional research questions to be answered.