`Ophthalmic Emulsion in the Treatment
`of Moderate-to-severe Dry Eye Disease
`A Dose-Ranging, Randomized Trial
`
`Dara Stevenson, MD,1 Joseph Tauber, MD,2 Brenda L. Reis, PhD,3 The Cyclosporin A Phase 2 Study Group
`
`Objective: To investigate the efficacy, safety, formulation tolerability, and optimal dosing of a novel cyclo-
`sporin A oil-in-water emulsion formulation for the treatment of moderate-to-severe dry eye disease.
`Design: Randomized, multicenter, double-masked, parallel-group, dose-response controlled trial.
`Participants: Total enrollment: 162 patients; cyclosporin A groups: 129 patients; vehicle group: 33 patients.
`Intervention: Patients instilled study medication (cyclosporin A ophthalmic emulsion 0.05%, 0.1%, 0.2%, or
`0.4%, or vehicle) twice daily into both eyes for 12 weeks, followed by a 4-week posttreatment observation period.
`Main Outcome Measures: Efficacy: rose bengal staining, superficial punctate keratitis, Schirmer tear test,
`symptoms of ocular discomfort, and the Ocular Surface Disease Index (OSDI; a measure of symptom frequency
`and impact on vision-related functioning). Safety: biomicroscopy, cyclosporin A blood levels, conjunctival
`microbiology, intraocular pressure, visual acuity, and monitoring of adverse events.
`Results:
`In a subset of 90 patients with moderate-to-severe keratoconjunctivitis sicca, the most significant
`improvements with cyclosporin A treatment were in rose bengal staining, superficial punctate keratitis, sandy or
`gritty feeling, dryness, and itching, with improvements persisting into the posttreatment period in some treatment
`groups. There was also a decrease in OSDI scores, indicating a decrease in the effect of ocular symptoms on
`patients’ daily lives. There was no clear dose-response relationship, but cyclosporin A 0.1% produced the most
`consistent improvement in objective and subjective end points and cyclosporin A 0.05% gave the most
`consistent improvement in patient symptoms. The vehicle also performed well, perhaps because of its long
`residence time on the ocular surface. There were no significant adverse effects, no microbial overgrowth, and no
`increased risk of ocular infection in any treatment group. The highest cyclosporin A blood concentration detected
`was 0.16 ng/ml. All treatments were well tolerated by patients.
`Conclusions: Cyclosporin A ophthalmic emulsions, 0.05%, 0.1%, 0.2%, and 0.4%, were safe and well
`tolerated, significantly improved the ocular signs and symptoms of moderate-to-severe dry eye disease, and
`decreased the effect of the disease on vision-related functioning. Cyclosporin A 0.05% and 0.1% were deemed
`the most appropriate formulations for future clinical studies because no additional benefits were observed with
`the higher concentrations. Ophthalmology 2000;107:967–974 © 2000 by the American Academy of Ophthal-
`mology.
`
`Recent population-based surveys indicate that dry eye dis-
`ease, or keratoconjunctivitis sicca, affects millions of people
`worldwide.1,2 Moreover, as many as 17%3 to 25%4 of
`
`Originally received: May 4, 1999.
`Manuscript no. 99216.
`Accepted: January 27, 2000.
`1Mercy Hospital Professional Building, New Orleans, Louisiana.
`2Kansas City, Missouri.
`3Allergan, Inc., Irvine, California.
`Reprint requests to Linda Lewis, 575 Anton Boulevard, Suite 900, Costa
`Mesa, CA 92626.
`Preliminary results of this study were presented at the annual meeting of
`the Association for Research in Vision and Ophthalmology, Fort Lauder-
`dale, Florida, May 1997, and at the XIth Congress of the European Society
`of Ophthalmology, Budapest, Hungary, June 1997.
`Supported by a grant from Allergan Inc.
`
`patients visiting ophthalmic clinics report dry eye symp-
`toms, making dry eye disease one of the most common
`complaints seen by ophthalmic specialists. Patients with dry
`eye disease typically complain of symptoms of ocular dis-
`comfort, including a dry, gritty feeling often accompanied
`by foreign body sensation. Depending on the duration and
`severity of disease, damage to the ocular surface may also
`be present. Patients with chronic, uncontrolled dry eye have
`an increased risk of ocular infections5,6 and are more likely
`to have ocular infections that progress to endophthalmitis.7
`A growing body of evidence suggests that chronic dry
`eye disease is the result of an underlying cytokine and
`receptor-mediated inflammatory process that affects the lac-
`rimal gland acini and ducts, leading to abnormalities in the
`tear film and ultimately disrupting the homeostasis of the
`ocular surface.8 –11 Most conventional treatments for dry
`
`© 2000 by the American Academy of Ophthalmology
`Published by Elsevier Science Inc.
`
`ISSN 0161-6420/00/$–see front matter
`PII S0161-6420(00)00035-X
`
`967
`
` EXHIBIT 1015
`
`
`
`Ophthalmology Volume 107, Number 5, May 2000
`
`eye disease focus on tear replacement or tear preservation
`and are incapable of affecting these processes. However,
`topical treatment with the immunomodulatory agent cyclo-
`sporin A has been shown to reduce cell-mediated inflam-
`matory responses associated with inflammatory ocular sur-
`face diseases.12,13 Preliminary studies have demonstrated
`that
`treatment with topical cyclosporin A can result
`in
`improvement of the signs and symptoms of dry eye disease
`(Foulks et al, Invest Ophthalmol Vis Sci 1996;37(Suppl):
`S646; Helms et al, Invest Ophthalmol Vis Sci 1996;37
`(Suppl):S646).12,14,15 In addition, several studies have es-
`tablished the efficacy of topical cyclosporin A in the treat-
`ment of keratoconjunctivitis sicca in dogs.16 –18 These find-
`ings suggest
`that
`topical cyclosporin A may provide a
`unique opportunity to move beyond treatments that only
`alleviate the symptoms of dry eye disease to therapies that
`effectively target the inflammatory processes contributing to
`disease pathogenesis.
`The purpose of this study was to investigate the efficacy,
`safety, patient tolerability, and optimal dosing of a novel
`cyclosporin A oil-in-water emulsion formulation for the
`treatment of moderate-to-severe dry eye disease with or
`without Sjo¨gren’s syndrome.
`
`Methods
`
`Study Protocol
`
`This report describes a randomized, multicenter, double-masked,
`parallel-group, dose-response study. The protocol was composed
`of three phases: a 2-week washout phase, a 12-week treatment
`phase, and a 4-week posttreatment phase. This study was con-
`ducted in compliance with the institutional review board regula-
`tions, informed consent regulations, sponsor and investigator ob-
`ligations, and the Declaration of Helsinki. Written informed
`consent was obtained from all patients before the initiation of any
`study medication or study-related procedure.
`Study Population. Patients were recruited between May 1995
`and February 1996 from nine clinical sites throughout the United
`States. Eligible patients were at least 21 years of age and had a
`diagnosis of keratoconjunctivitis sicca with or without Sjo¨gren’s
`syndrome refractory to conventional management. Inclusion crite-
`ria included Schirmer test (without anesthesia) of 7 mm/5 minutes
`in at least one eye; mild superficial punctate keratitis defined as a
`corneal punctate fluorescein staining score of $ 1 in either eye
`(scale 0 [none] to 3 [severe]); and one or more moderate ($12)
`dry eye–related symptoms, including itching, burning, blurred
`vision, foreign body sensation, dryness, photophobia, and soreness
`or pain. Both eyes were treated, but both eyes were not included in
`all analyses (see Statistical Methods).
`Patients were excluded from study participation if they had any
`ocular disorder including ocular injury, infection, non-dry eye
`ocular inflammation, trauma, or surgery within the prior 6 months;
`were receiving concurrent treatment that could interfere with in-
`terpretation of the study results; had any uncontrolled systemic
`disease or significant illness; or were pregnant, lactating, or con-
`sidering a pregnancy.
`Study Medications. The medications used in this study were
`unit dose vials of unpreserved cyclosporin A 0.05%, 0.1%, 0.2%,
`and 0.4% ophthalmic emulsion; unit dose vials of unpreserved
`vehicle for cyclosporin A 0.2% ophthalmic emulsion; and RE-
`FRESH lubricant eye drops (Allergan, Irvine, CA). The vehicle for
`
`968
`
`each concentration of cyclosporin A ophthalmic emulsion is for-
`mulated slightly differently because greater oil content is required
`to dissolve the higher concentrations of the active ingredient. The
`vehicle for cyclosporin A 0.2% ophthalmic emulsion (hereafter
`referred to as “vehicle”) was chosen for the control because it was
`near the middle of the range of cyclosporin A concentrations used.
`Study Treatments. During the washout phase, patients were
`instructed to discontinue use of all topical ophthalmic medications
`except for REFRESH. During this time, they were instructed to use
`REFRESH a minimum of four but no more than eight times daily
`in each eye. Patients who successfully completed the washout
`phase were then given their assigned medication (cyclosporin A
`0.05%, 0.1%, 0.2%, or 0.4% ophthalmic emulsion or emulsion
`vehicle) and instructed to instill
`their medication twice daily
`(morning and evening) in both eyes for 12 weeks. The use of
`REFRESH (up to eight times daily in each eye) was allowed
`during the treatment phase.
`Outcome Measures. The efficacy measures were rose bengal
`staining (graded on a scale from 0 5 none to 3 5 severe);
`superficial punctate keratitis measured at nasal, temporal, pupil,
`and inferior and the scores summed (each graded on a scale from
`0 5 none to 3 5 severe); Schirmer tear test (without anesthesia,
`with nasal stimulation only if needed to determine that the patient
`had some capacity to secrete tears); symptoms of ocular discom-
`fort (graded by investigator queries on a scale from 0 5 none to
`4 5 very severe, and in patient diaries on a scale from 0 5 no
`discomfort to 4 5 discomfort that interferes with normal daily
`activity); tear film debris (graded on a scale of 0 5 none to 4 5
`very severe); tear breakup time; and the frequency and amount of
`REFRESH used.
`In addition, patient response to treatment was evaluated using
`the Ocular Surface Disease Index (OSDI), a global assessment
`parameter consisting of 12 questions designed to assess the symp-
`toms of ocular irritation consistent with dry eye disease and their
`impact on vision-related functioning. The questions covered three
`areas: ocular symptoms, environmental triggers, and vision-related
`function. Each question was phrased in terms of frequency (how
`often they were aware of a symptom, how often they experienced
`difficulty with a specific task because of their symptoms, etc) and
`graded on a scale from 0 to 4 (where 0 5 “never” and 4 5 “all the
`time”). Patient responses to all answers were then combined for a
`composite OSDI score ranging from 0 to 100.
`Treatment safety was assessed by biomicroscopy, measurement
`of cyclosporin A blood levels, conjunctival microbiology, hema-
`tology and blood chemistry panels, intraocular pressure by appla-
`nation tonometry, and visual acuity by a 96% contrast Regan
`Letter Acuity Chart. Throughout the study, patients were moni-
`tored for signs and symptoms of adverse events and formulation
`tolerability. Any reported adverse event was graded by the inves-
`tigator for severity (mild, moderate, or severe) and assessed for
`relationship to the study treatment (none, unlikely, possible, prob-
`able, definite, or unknown).
`Patients were evaluated at weeks 4, 8, and 12 during the
`treatment phase. During these visits patients were evaluated for
`changes from baseline in Schirmer tear test, rose bengal staining,
`superficial punctate keratitis scores, symptoms of ocular discom-
`fort, biomicroscopy, and visual acuity. After the completion of the
`treatment phase, patients were also evaluated at posttreatment
`weeks 2 and 4. During both visits patients were assessed for Schirmer
`tear test, rose bengal staining, superficial punctate keratitis, ocular
`symptoms of discomfort, biomicroscopy, and visual acuity.
`Whole blood was obtained from all patients for evaluation of
`cyclosporin A trough levels at baseline; treatment weeks 1, 4, and
`12; and posttreatment week 4. At week 12, additional blood
`samples were drawn at one study site only for evaluation of peak
`cyclosporin A concentrations. For evaluation of trough levels,
`
`
`
`Stevenson et al
`
`z Cyclosporin A in Treatment of Dry Eye Disease
`
`blood was drawn immediately before the morning dose of study
`medication. For evaluation of peak levels, blood was drawn 1, 2,
`and 4 hours after instillation of the final dose of study medication
`at week 12.
`Blood samples were sent to the Allergan Pharmacokinetics
`Laboratory, where they were assayed by liquid chromatography-
`mass spectroscopy/mass spectroscopy (LC-MS/MS) with a detec-
`tion limit of 0.1 ng/ml. One milliliter of human blood was acidified
`with 2 ml of 0.1 N HCl solution and analytes extracted with 5 ml
`of methyl t-butyl ether. After separation from the acidified aqueous
`layer, the organic layer was made basic with 2 ml of 0.1 N NaOH,
`centrifuged, and the organic extract was then evaporated. The
`dried extract was reconstituted in 200 ml of mobile phase A and B
`(1:1 v/v) and 100 ml was injected into the LC-MS/MS system. The
`LC-MS/MS analysis was conducted on a PE-Sciex API III1 triple
`quadrupole mass spectrometer (Perkin Elmer, Norwalk, CT) cou-
`pled to a Shimadzu HPLC system (Columbia, MD). Chromatog-
`raphy was performed on a Keystone BDS Hypersil C8 column
`(50 3 2 mm, 3 mm) with a binary mixture of 2 ammonium
`acetate/0.4% formic acid in water (mobile phase A) and 2 mmol/l
`ammonium acetate/0.4% formic acid in acetonitrile (mobile phase
`B) under gradient elution. The HPLC effluent flow rate of 300
`ml/min was split, with 75 ml/min directed to the atmospheric
`ionization source. The mobile phase was 60% B at 0 to 0.5 minute,
`increased linearly to 95% B at 1 minute, held at 95% B from 1 to
`2.5 minutes, and then decreased to 60% B at 3 minutes (held 1
`minute). Cyclosporin G was used as the internal standard.
`The PE-Sciex MacQuan software (PE-Sciex Instruments, Con-
`cord, Ontario, Canada) was used to determine peak areas of
`analyte and internal standard, peak area ratios of analyte/internal
`standard, calibration curves, and calculated concentrations of un-
`knowns. The accuracy and precision of the LC-MS/MS method
`was assessed within each run using quality control blood samples
`at 0.1, 0.2, 1, and 5 ng/ml. The intraday accuracy (percent ratio of
`observed to nominal concentration) ranged from 100% to 109%,
`with precision (coefficient of variation) ranging from 3% to 10%.
`The interday accuracy ranged from 102% to 113%, with precision
`ranging between 1% and 13%.
`At four selected study centers, ocular samples for microbio-
`logic evaluation were collected from the conjunctival cul de sac at
`baseline, treatment week 12, and posttreatment week 4 and sent to
`a centralized laboratory for culture and organism identification.
`Statistical Methods. Efficacy variables from subjective mea-
`surements with data collected on both eyes were analyzed by
`averaging the data from both eyes. Efficacy variables from objec-
`tive measurements with data collected on both eyes were analyzed
`using data from the worse eye. The worse eye was defined as the
`eye with the worse Schirmer value and the worse superficial
`punctate keratitis value (pupil and nasal areas only) at baseline.
`Because of the heterogeneity of patient disease profiles, sub-
`group analyses of patients who had various degrees of disease
`severity were analyzed separately. Only patients with moderate-
`to-severe dry eye disease at baseline were included in the efficacy
`analysis described in this report. All patients who received study
`medication were included in the analysis of safety variables.
`Demographic parameters were summarized with descriptive
`statistics and frequency tables. Efficacy parameter comparisons
`among treatment groups were analyzed with the Kruskal-Wallis
`test. Pairwise comparisons between treatment groups were ana-
`lyzed with the Wilcoxon rank sum test. Within-group changes
`from baseline were evaluated with the Wilcoxon signed-rank test.
`REFRESH use, intraocular pressure, and laboratory variables were
`analyzed by analysis of variance. Within-group changes from
`baseline were evaluated with the paired t test. Adverse event data
`were summarized by frequency tables. A two-sided test with P 5
`0.05 was considered statistically significant for all main effects.
`
`The null hypothesis was that there were no differences among
`the treatment groups with regard to changes from baseline values.
`The alternative hypothesis was that there was a change.
`Power was calculated to detect an among-group difference in
`change from baseline in categorized Schirmer tear values at week
`12. For a sample size of 12 to 15 patients in the moderate-to-severe
`subgroup analysis, a standard error of 0.394 and standard deviation
`of 0.881, the power to detect a one grade difference was 0.69.
`
`Patient Treatment Assignment
`Qualified patients within each investigator’s population were as-
`signed equally to one of the five masked treatment groups sequen-
`tially, corresponding to a randomization schedule generated by the
`sponsor and using a block of five design.
`
`Study Masking
`All study medications were liquids of similar appearance, dis-
`pensed in identical unit dose vials, sealed in identical two-com-
`partment plastic pouches, and packed in identical boxes of 16
`pouches each. Each pouch and packing box was coded with a
`shipment number and the patient number.
`When each box was dispensed, the tear-off portion of the label
`was attached to the patient’s case report form. If necessary (be-
`cause of a serious or severe adverse event), the investigator could
`irreversibly unmask the tear-off portion of the patient’s medication
`label to determine which study medication the patient had received
`to institute appropriate patient care.
`
`Results
`
`Because this was the first clinical trial conducted with this new
`cyclosporin A formulation, it was designed to function as a pilot
`study for future investigations. Therefore, patients who varied
`widely in the severity of their dry eye were enrolled. The data from
`all patients who received study medication (intent-to-treat popu-
`lation) were analyzed. However, a subgroup analysis revealed a
`sizable population of patients who had moderate-to-severe dry eye
`disease at baseline. Moderate-to-severe dry eye disease was de-
`fined as a Schirmer tear test 5 mm/5 minutes at baseline in at least
`one eye and superficial punctate keratitis (pupil and nasal average)
`of 1.5 averaged over both eyes. Because these patients represent
`the greatest therapeutic challenge for any dry eye treatment, the
`efficacy analysis presented here is confined to the evaluation of
`this moderate-to-severe subgroup. This subgroup also represents
`the most appropriate target population for future clinical studies of
`dry eye therapeutics because these patients have sufficient mani-
`festations of the disease to allow the response to therapeutic
`intervention to be more objectively evaluated. Data from all pa-
`tients were included in the safety analysis.
`
`Participant Flow and Follow-up
`A total of 162 patients was enrolled: 129 in the cyclosporin A
`groups and 33 in the vehicle group (Table 1). Eight patients were
`discontinued for administrative reasons. Of the four patients dis-
`continued because of adverse events, two were in the vehicle group
`(one with a visual disturbance and ocular burning, and one with
`conjunctivitis and a contact irritation dermatitis), one was in the
`cyclosporin A 0.2% group (ocular burning) and one was in the
`cyclosporin A 0.4% group (myocardial infarction). Only the ocular
`adverse events were considered to be possibly or probably related
`to the study medication.
`
`969
`
`
`
`Ophthalmology Volume 107, Number 5, May 2000
`
`Table 1. Patient Disposition
`
`Moderate-to-Severe Dry Eye Disease
`(n 5 90)
`
`Completed
`
`%
`
`Discontinued
`n
`%
`
`100.0
`100.0
`94.7
`100.0
`94.4
`
`97.8
`
`0
`0
`1
`0
`1
`
`2
`
`0.0
`0.0
`5.3
`0.0
`5.6
`
`2.2
`
`n
`
`16
`17
`18
`20
`17
`
`88
`
`n
`
`30
`30
`30
`32
`28
`
`150
`
`Treatment
`Group
`
`Vehicle
`CsA 0.05%
`CsA 0.1%
`CsA 0.2%
`CsA 0.4%
`
`Total
`
`Intent-to-Treat Population
`(Total Enrollment) (n 5 162)
`Completed
`Discontinued
`%
`
`%
`
`n
`
`90.9
`96.8
`93.8
`94.1
`87.5
`
`92.6
`
`3
`1
`2
`2
`4
`
`12
`
`9.1
`3.2
`6.3
`5.9
`12.5
`
`7.4
`
`CsA 5 cyclosporin A.
`
`Of the 90 patients with moderate-to-severe dry eye disease, 16
`were in the vehicle group, 17 in the cyclosporin A 0.05% group, 19
`in the cyclosporin A 0.1% group, 20 in the cyclosporin A 0.2%
`group, and 18 in the cyclosporin A 0.4% group. One patient in the
`cyclosporin A 0.1% group was discontinued for personal reasons,
`and one patient in the cyclosporin A 0.4% group was discontinued
`because of a myocardial infarction (same patient as mentioned
`earlier). No patients’ medications were unmasked during this
`study.
`
`Patient Demographics
`The demographic characteristics of the patient population are
`listed in Table 2. Note that the mean patient age was approximately
`58 years, that more than 80% of patients were women, and that
`approximately 90% were white. Approximately 32% of the pa-
`tients in the moderate-to-severe dry eye group were also Sjo¨gren’s
`syndrome patients. Sjo¨gren’s syndrome was defined as the pres-
`ence of one or more of the following in the blood: antinuclear
`antibodies (.0 titer); rheumatoid factor (30 international units/
`ml); Sjo¨gren’s syndrome A (.10 IU/ml) or B (.5 IU/ml) anti-
`bodies. No significant differences were noted among the treatment
`groups for either the intent-to-treat or moderate-to-severe dry eye
`populations.
`
`Efficacy Analysis
`At baseline, mean scores for conjunctival rose bengal staining
`ranged from 1.2 to 2.0 for both temporal and nasal regions in all
`
`Table 2. Patient Demographics
`
`Age
`Mean (range)
`Gender (%)
`Male
`Female
`Race (%)
`White
`Black
`Asian
`Hispanic
`Sjo¨gren’s syndrome (%)
`
`Moderate-to-Severe
`Dry Eye Disease
`(n 5 90)
`
`Intent-to-Treat
`Population
`(n 5 162)
`
`58 (31–88)
`
`59 (31–88)
`
`17 (18.9)
`73 (81.1)
`
`82 (91.1)
`4
`(4.4)
`0
`(0.0)
`4
`(4.4)
`29 (32.2)
`
`26 (16.0)
`136 (84.0)
`
`145 (89.5)
`12
`(7.4)
`1
`(0.6)
`4
`(2.5)
`43 (26.5)
`
`CsA 5 cyclosporin A; NA 5 data not available.
`
`970
`
`treatment groups. Significant improvements from baseline in tem-
`poral conjunctival rose bengal staining scores were observed with
`cyclosporin A 0.1% at all treatment and posttreatment visits (P #
`0.016), with cyclosporin A 0.2% at week 12 and both posttreat-
`ment visits (P # 0.047), with cyclosporin A 0.4% at week 8 (P 5
`0.031), and with the emulsion vehicle at week 12 (P 5 0.047) (Fig
`1). Cyclosporin A 0.1% produced significantly greater improve-
`ments in temporal conjunctival rose bengal staining scores than
`vehicle (P 5 0.006), cyclosporin A 0.05% (P 5 0.022), and
`cyclosporin A 0.4% (P 5 0.007) at posttreatment week 2.
`Significant improvements from baseline in nasal conjunctival
`rose bengal staining scores were observed with cyclosporin A
`0.2% at all treatment and posttreatment visits (P # 0.022), with
`cyclosporin A 0.1% and 0.05% at treatment week 4 through
`posttreatment week 2 (P # 0.031), in the cyclosporin A 0.4%
`group at posttreatment week 2 (P 5 0.031), and in the vehicle
`group at treatment weeks 8 and 12 (P # 0.025). There were no
`significant among-group differences in the change from baseline in
`nasal conjunctival rose bengal staining.
`At baseline, mean scores for superficial punctate keratitis
`ranged from 1.6 to 1.9 in all treatment groups. Cyclosporin A 0.1%
`produced the greatest improvement from baseline in superficial
`punctate keratitis scores throughout the treatment and posttreat-
`ment periods (range, 20.9 to 21.4 units) (Fig 2). With the excep-
`tion of the 0.05% concentration at treatment week 12 and post-
`treatment week 4, significant
`improvements from baseline in
`superficial punctate keratitis were seen in all cyclosporin A treat-
`
`Figure 1. Change from baseline in temporal rose bengal staining. CsA,
`Cyclosporin A. *, Significantly different from baseline (P # 0.047); †,
`Significantly different from vehicle, cyclosporin A 0.05%, and cyclosporin
`A 0.4% (P # 0.022).
`
`
`
`Stevenson et al
`
`z Cyclosporin A in Treatment of Dry Eye Disease
`
`Figure 2. Change from baseline in superficial punctate keratitis. CsA,
`Cyclosporin A. *, Significantly different from baseline (P # 0.018).
`
`ment groups at all time points during the 12-week treatment phase
`(P # 0.012) and 4-week posttreatment period (P # 0.018). Sig-
`nificant improvement in superficial punctate keratitis was also
`observed in patients treated with vehicle at treatment weeks 8 and
`12 and posttreatment week 2 (P # 0.041). No statistically signif-
`icant among-group differences in superficial punctate keratitis
`values were observed.
`Baseline values for Schirmer tear test wetting scores ranged
`from 2.4 to 3.1 in all treatment groups. The most consistent
`improvements were in the cyclosporin A 0.1% group, with mean
`increases in wetting length of 4.3 mm at week 8 and 2.8 mm at
`week 12, but these increases only approached statistical signifi-
`cance (week 8, P 5 0.051; week 12, P 5 0.055). The only
`statistically significant improvement from baseline occurred in the
`cyclosporin A 0.4% group at treatment week 4 (P 5 0.008) and
`posttreatment week 4 (P 5 0.023), whereas a significant worsen-
`ing occurred in the vehicle group at week 4 (3.0 mm, P 5 0.047).
`Cyclosporin A 0.4% produced significantly (P # 0.025) greater
`improvements from baseline in Schirmer tear test results than
`either vehicle or cyclosporin A 0.2% at posttreatment week 4.
`Symptoms of ocular discomfort were evaluated from scheduled
`visit queries from the clinical investigator and from self-adminis-
`tered, weekly patient diaries. Baseline symptom results suggest
`that patients may have consistently underreported the severity of
`their symptoms when responding to the query from the health
`professional compared with what they recorded in their diaries.
`Therefore, only symptom data from patient diaries (using the
`entries immediately before each scheduled visit) are presented.
`At baseline, the mean score for sandy or gritty feeling ranged
`from 1.7 to 2.2 (mild to moderate) in all treatment groups. There
`was a significant improvement from baseline in sandy or gritty
`feeling in the cyclosporin A 0.05% and 0.4% groups at several
`visits (P # 0.039) (Fig 3). At treatment week 12, all cyclosporin
`A treatment groups had significantly greater improvements in
`sandy or gritty feeling than the vehicle group (P # 0.04). This
`significant difference from vehicle was also seen at posttreatment
`week 2 in the cyclosporin A 0.05% and 0.4% groups (P # 0.006)
`and at posttreatment week 4 in the cyclosporin A 0.05%, 0.2%, and
`0.4% groups (P # 0.027). At posttreatment week 2, the cyclo-
`sporin A 0.4% and 0.05% groups also demonstrated a significantly
`greater improvement than the cyclosporin A 0.2% group (P #
`0.037).
`At baseline, the mean score for ocular dryness ranged from 2.3
`to 2.7 (moderate to severe) in all treatment groups. Significant
`improvements from baseline in ocular dryness were seen at two or
`more time points in all cyclosporin A groups except the cyclo-
`sporin A 0.1% group (P # 0.036) (Fig 4). At posttreatment week
`
`Figure 3. Change from baseline in sandy or gritty feeling. CsA, Cyclo-
`sporin A. *, Significantly different from baseline (P # 0.039); †, signifi-
`cantly different from vehicle (P # 0.027); §, significantly different from
`cyclosporin A 0.2% (P # 0.037).
`
`4, cyclosporin A 0.05%, 0.2%, and 0.4% groups all demonstrated
`significantly greater improvements in ocular dryness than did the
`vehicle group (P # 0.010).
`At baseline, the mean score for ocular itching ranged from 1.4
`to 1.9 (mild to moderate) in all treatment groups. Significant
`improvements from baseline in ocular itching were seen at one or
`more time points in all of the cyclosporin A groups (P # 0.031)
`but not in the vehicle group. The magnitude of improvement in the
`cyclosporin A groups was larger than that in the vehicle group at
`all time points, but there were no statistically significant differ-
`ences among any of the groups at any time point.
`There were no significant within-group or between-group dif-
`ferences in photophobia, pain, or burning and stinging at any time
`point. The mean scores at baseline for all these parameters ranged
`from 1 to 2 (mild to moderate) in all treatment groups.
`Baseline OSDI scores ranged from 33 to 42 (on a scale from 0
`to 100, where 0 indicates no disability and 100 indicates complete
`disability) in all treatment groups. At both treatment week 12 and
`posttreatment week 4, there was at least a trend toward improve-
`ment in the OSDI score in the cyclosporin A 0.1%, 0.2%, and 0.4%
`groups, whereas there was either no change or worsening in the
`vehicle group (Fig 5). At week 12, cyclosporin A 0.1% and 0.2%
`significantly reduced OSDI scores (P # 0.008). This decrease was
`significantly greater with cyclosporin A 0.1% than with cyclo-
`sporin A 0.05% or 0.2% (P # 0.032). This improvement in OSDI
`scores persisted into the posttreatment period, with a significant
`
`Figure 4. Change from baseline in ocular dryness. CsA, Cyclosporin A. *,
`Significantly different from baseline (P # 0.036); †, significantly different
`from vehicle (P # 0.010).
`
`971
`
`
`
`Ophthalmology Volume 107, Number 5, May 2000
`
`Table 3. Change in Ocular Microflora From Baseline to Week 12
`
`Treatment
`Group
`
`Vehicle
`CsA 0.05%
`CsA 0.1%
`CsA 0.2%
`CsA 0.4%
`
`n
`
`14
`15
`14
`15
`16
`
`Number of Patients
`
`No Growth to
`Any Growth
`
`Any Growth to
`No Growth
`
`Change in
`Gram’s Stain
`
`4
`3
`3
`1
`5
`
`1
`3
`2
`4
`0
`
`1
`0
`1
`1
`2
`
`CsA 5 cyclosporin A.
`No growth to any growth 5 no bacteria at baseline but the presence of
`bacteria at week 12.
`Any growth to no growth 5 presence of bacteria at baseline but no
`bacteria at week 12.
`Change in Gram’s stain 5 from gram-negative to gram-positive, or
`conversely.
`
`A total of 15 treatment-related adverse events were reported by
`the 162 patients in the intent-to-treat group (Table 4). The greatest
`incidence of treatment-related adverse events occurred in patients
`in the vehicle group (3 of 33; 9.1%), and the most frequently
`reported adverse event was superficial punctate keratitis (5 of 162;
`3%). No deaths or serious treatment-related adverse events oc-
`curred in any patient during the study.
`No clinically significant changes in visual acuity or blood
`chemistry and hematology values (including liver and renal func-
`tion parameters) were observed in any treatment group, and no
`patients experienced adverse events related to blood chemistry or
`hematology. Only one patient had a clinically significant increase
`in intraocular pressure from baseline (defined as an intraocular
`pressure that was within the normal range at baseline and increased
`to .25 mmHg) and that patient was in the vehicle group. There
`were no statistically or clinically significant within-group or
`among-group changes in any biomicroscopic variables, except at
`week 8 in the vehicle group in which there was a statistically
`significant increase from baseline in lid erythema (P 5 0.016).
`
`Discussion
`
`The most important results of this study were that treatment
`with topical cyclosporin A 0.05% to 0.4% ophthalmic emul-
`sions significantly improved the ocular signs and symptoms
`of moderate-to-severe dry eye disease and that these im-
`provements resulted in a significant decrease in the effect of
`the disease on vision-related functioning (as measured by
`the OSDI).
`These findings support the results of earlier studies that
`demonstrated a beneficial effect of topical cyclosporin A on
`dry eye disease15 or dry eye in Sjo¨gren’s syndrome14
`(Foulks et al, Invest Ophthalmol Vis Sci 1996;37(Suppl):
`S646; Helms et al, Invest Ophthalmol Vis Sci 1996;37
`(Suppl):S646) and expands the range of patients who may
`potentially benefit from cyclosporin A treatment to those
`with moderate-to-severe dry eye disease with or without
`Sjo¨gren’s syndrome. In addition, these findings may provide
`further insight into the pathophysiology of dry eye disease.
`The mechanisms contributing to dry eye disease are still
`being explained, but several lines of evidence suggest that
`
`Figure 5. Change from baseline in the OSDI. CsA, Cyclosporin A. *,
`Significantly different from baseline (P # 0.026); †, significantly different
`from cyclosporin A 0.05% and cyclosporin A 0.2% (P # 0.032).
`
`improvement seen in the cyclosporin A 0.2% group (P 5 0.026)
`and a trend toward significance seen in the cyclosporin A 0.1%
`group (P 5 0.055).
`Only cyclosporin A 0.1% produced significant improvements
`from baseline values in tear film debris at any time point (week 8,
`P 5 0.005; week 12, P 5 0.031). There were no significant
`differences in tear break up time among the treatment groups.
`At treatment week 12 and both posttreatment visits, patients in
`the vehicle group used a greater number of REFRESH units per
`day (mean values, week 12, 4.3 units; post