Abstract
Purpose
To evaluate the efficacy and safety of a preservative-free cationic emulsion (CE) with a 0.18% hyaluronate sodium (HS) solution in patients with moderate to severe dry eye disease (DED) with keratitis or keratoconjunctivitis.
Methods
Eighty-five patients were randomized (1:1) in this multicenter, prospective, reference-controlled, parallel-group, investigator-masked study to receive CE (n = 44) or HS (n = 41). Clinical signs and symptoms were assessed over 3 months. The primary efficacy endpoint was noninferiority of CE to HS in change from baseline of ocular surface staining (OSS) score at 1 month.
Results
In the per protocol (PP) set and full analysis set (FAS), CE showed a similar and noninferior (p<0.0001) improvement in OSS scores compared with HS at 1 month (PP: −2.5 ± 1.3 vs −1.9 ± 1.6; FAS: −2.2 ± 1.5 vs −2.0 ± 1.8 for CE vs HS). Other clinical signs of DED similarly improved in both groups. In the FAS, global symptoms score of ocular discomfort was significantly better with CE compared with HS at 1 month (-14.8 ± 17.3 vs −7.6 ± 14.2; p = 0.0469), including greater alleviation of itching (-14.8 ± 21.2 vs −1.7 ± 19.7; p = 0.0100) and eye dryness (-21.9 ± 28.3 vs −8.4 ± 21.4; p = 0.0016). Similar trends were observed at 3 months for itching and eye dryness. Investigator global efficacy assessment and quality of life scores for eye pain and driving favored CE at 3 months. Incidence of adverse events was low in both treatment groups.
Conclusions
CE was similar to HS with regards to safety and efficacy for objective signs but was superior to HS in improving DED symptoms in patients with moderate to severe DED.
Introduction
Dry eye disease (DED) is a complex ocular disease that can involve both the lacrimal system and the ocular surface (1). Dry eye disease can affect up to 34% of certain populations, particularly female and elderly (>65 years old) patients (2, 3). The chronic nature of DED means that this common condition can significantly impact patient quality of life and impose a significant cost burden on society (4, 5).
DED has traditionally been classified clinically into one of 2 major subtypes: aqueous tear deficiency or evaporative tear dysfunction (6). In reality, however, DED pathology is often not discretely organized but rather is a continuous cascade of interacting mechanisms, all of which can interact to drive a vicious circle of increasing inflammation and disease severity, if not adequately treated (7).
The classic tear film model is composed of 3 major components: an outer lipid layer, an aqueous middle layer, and the inner mucus layer. Intrinsic or extrinsic changes to any of these 3 tear film components can affect ocular surface homeostasis and encourage tear film instability and tear hyperosmolarity (1, 8-11).
There is no gold standard treatment for DED (12). In clinical practice, first-line treatment usually focuses on achieving adequate relief of DED symptoms, traditionally with pharmacologically inactive aqueous solution- or aqueous gel-based artificial tears (13, 14). These solutions replenish the aqueous phase of the tear film and have been shown to work well in relieving symptoms of DED (15, 16). However, their effect can be relatively transient, and the efficacy of these drops on moderate to severe DED may not be sufficient (17). This is partly because reflux tearing and drainage via the nasolacrimal duct clears ≥80% of the instilled drug very quickly after application (18). Furthermore, most aqueous eyedrops lack lipids in their composition (19).
More recent research has focused on creating artificial tear formulations that better mimic tears or tear film composition, including lipids, with reduced clearance and improved efficacy or tolerability. A range of ocular lubricants are now available with differing electrolyte balance, osmolarity (i.e., by addition of compatible solutes and osmoprotectants such as glycerin) (20), and viscosity (through addition of macromolecular complexes such as carboxymethyl cellulose (CMC), polyvinyl alcohol, oils, or hyaluronic acid), and which contain less toxic preservatives (i.e., sodium chlorite) or are preservative-free (13).
Novel oil-based or lipid-containing compounds may help to restore tear film stability and potentially provide a method of ocular drug delivery (19, 21). Emulsions are a relatively new and promising type of ophthalmic formulation (22-25). They are essentially composed of oil nanodroplets stabilized by interfacial surfactants, dispersed in an aqueous phase (14). Cationic (i.e., positively charged) emulsions are a new type of emulsion developed as a treatment option for DED (19). It is thought that the positive charge of the nanodroplets enables them to adhere to the negatively charged cells of the ocular surface, creating a lubricating and sustainable lipid layer (19). The nanodroplets may also help to replenish tear film layers and restore stability of the tear film (14, 21, 26). It is therefore proposed that the use of cationic emulsions (CEs) in the treatment of DED may afford longer-lasting hydration and lubrication of the ocular surface when compared with other artificial tear products.
Cationorm® is a preservative-free CE presented in a sterile single-dose container. The safety and efficacy of this emulsion for ocular use have been established in previous clinical studies in DED (27, 28).
The objective of the present study was to assess the ocular efficacy, tolerance, and safety of CE in comparison with reference treatment—0.18% hyaluronate sodium (HS) solution—in patients with bilateral moderate to severe DED with keratitis or keratoconjunctivitis.
Methods
Study design
This was a prospective, multicenter, randomized, investigator-masked, parallel-group, reference-controlled, noninferiority, 3-month study conducted in patients with moderate to severe DED with keratitis or keratoconjunctivitis. The trial was conducted at 15 sites in France in compliance with Good Clinical Practice and the principles outlined in the latest version of the Declaration of Helsinki. The first patient was randomized on October 28, 2011, and the last patient visit was on March 19, 2013. This article complies with CONSORT 2010 guidelines for reporting clinical study results. This study was performed with informed consent and following all the guidelines for experimental investigations required by the Ethics Committee.
This study was registered under the following number in the EudraCT database: 2011-A00955-36 with the protocol code number NVG11F120.
Patients
Patients were men or women ≥18 years of age who had been using or had needed to use artificial tears (preservative-free) for the treatment of DED for at least 3 months. Patients were required to have been experiencing at least 2 symptoms of ocular discomfort (e.g., itching, eye dryness, sticky feeling, photophobia, pain, burning or stinging, sandy feeling or grittiness, or foreign body sensation) with a score of ≥23 mm on an ocular discomfort visual analogue scale (VAS; 0% to 100% discomfort).
Patients must also have had (in at least one eye) ocular surface staining (OSS; the sum of nasal and temporal interpalpebral conjunctival and corneal vital staining using 2% fluorescein and lissamine green) with a modified Oxford scale score of ≥4 and ≤9 and a tear break-up time (TBUT; sum of 3 measurements) of ≤30 seconds or Schirmer tear test (without anesthesia) score of ≥3 and ≤9 mm/5 min.
Patients were excluded from the study if they had a corneal fluorescein staining (CFS) score ≥4 on a modified Oxford scale; ocular hypertension or glaucoma requiring treatment with an intraocular pressure–lowering medication; history of ocular trauma, infection, or ocular inflammatory condition not related to DED within the last 3 months of screening; moderate or severe blepharitis and/or meibomian gland disease; filamentary keratitis; any ocular surface anomaly not related to DED; or refractive surgery (e.g., laser-assisted in situ keratomileusis, laser-assisted subepithelial keratomileusis, photorefractive keratectomy) within 6 months and/or any other ocular laser/surgery within 3 months before the screening visit.
Study schedules
Patients signed an informed consent form at the screening visit (day -14 to day -7), and eligibility, concomitant therapies, medical history, and demographic characteristics were recorded. Each patient was subject to a 1- to 2-week washout period prior to randomization (from day -14 or day -7 to day 0 [baseline]) using Unilarm® artificial tears (Laboratoires Europhta, Monaco) as needed.
At baseline, patients were randomized 1:1, balanced by centers, to receive either hypotonic CE (Cationorm® preservative-free single-dose eye drops; Santen SAS, Evry, France) or reference treatment of hypotonic 0.18% HS solution (Vismed® preservative-free single-dose eye drops; TRB Chemedica, Geneva, Switzerland). Treatments were allocated according to a sequential randomization list, by blocks, prepared in advance by the clinical supplies distributor LC2 (ZA du Charpenay, Lentilly, France).
The study was conducted in a single-masked fashion with the investigator (who distributed the treatments) masked to treatment allocation. Test and reference medications were packaged similarly.
Patients were instructed to use 1 drop of study treatment in each eye 4 times daily (morning, noon, afternoon, and after dinner) from day 0 (baseline) to day 84 (± 3 days).
Three study visits were planned during the treatment period: at 1 week (day 7), 1 month (day 28 ± 3 days), and 3 months (day 84 ± 7 days).
Objective assessments of clinical signs
Corneal fluorescein staining followed by lissamine green staining of the interpalpebral regions of the nasal and temporal conjunctiva was assessed at screening and at all study visits. Staining was assessed with the slit-lamp using appropriate filters from 1 to 4 minutes after instillation of the respective dyes and graded using the modified Oxford scale (29).
Other clinical signs assessed using standard methods included TBUT (sum of 3 assessments within 1 minute of dye instillation) at baseline, day 28, and day 84 and Schirmer tear test without anesthesia (score at 5 minutes, at least 15 minutes after CFS and lissamine green staining) at baseline and day 84.
For exploratory purposes, tear film osmolarity was assessed at baseline, day 28, and day 84 only in those centers that had access to the TearLab Osmolarity System® (TearLab, San Diego, CA, USA). Selected sites also performed conjunctival impression cytology according to standard methods at baseline only, to investigate the relationship between human leukocyte antigen-DR (HLA-DR) expression (a marker of ocular surface inflammation [30]) and the demographic characteristics of the patients (31).
Subjective assessments
Symptoms of ocular discomfort (unrelated to instillation of study treatment) were assessed at screening and all study visits. Symptoms were assessed by the patients using a self-administered VAS ranging from 0% to 100% based on established DED symptomatology (itching, eye dryness, sticky feeling, photophobia, pain, burning or stinging, sandy feeling or grittiness, or foreign body sensation). The means of these 7 VAS scores were then combined into a global symptoms score.
The investigator at each center also rated overall efficacy of the study treatment at day 7, day 28, and day 84 on a 4-point scale ranging from 0 (unsatisfactory) to 3 (very satisfactory).
Patient quality of life was assessed using the 25-item National Eye Institute Visual Function Questionnaire (NEI VFQ-25) at baseline and day 84.
Safety assessments
Adverse events (AEs) were recorded from baseline to day 84. Best-corrected visual acuity and local ocular tolerance were assessed at screening and each study visit. Intraocular pressure was also measured (one measurement in each eye, the right preceding the left) using Goldmann applanation tonometry at baseline and day 84, after completion of all other ocular examinations.
Statistical analysis
Statistical analysis was carried out using SAS® software (SAS Institute, Cary, NC, USA). Efficacy evaluations were made only in the worst eligible eye according to OSS score or Schirmer test score (if no OSS difference) at baseline. The right eye was used if no difference in either of these scores was observed at baseline. All efficacy analyses were carried out on the full analysis set (FAS) and the per protocol (PP) set.
Primary efficacy endpoint analysis
The primary efficacy endpoint was predefined as OSS score at day 28. The primary efficacy analysis model was analysis of covariance (ANCOVA), performed to assess the estimate of the least squares means difference between groups in the change from baseline of the OSS score at day 28. The 95% confidence interval of this difference was computed and the upper boundary was compared with the noninferiority margin of 2 points. If the boundary was found to be lower than the noninferiority margin (in both the FAS and PP set), the noninferiority of CE to HS was assumed to be demonstrated.
Secondary efficacy endpoint analyses
Secondary efficacy endpoints of CFS and lissamine green staining (i.e., individual components of the ocular OSS) were analyzed according to the same ANCOVA model as the primary efficacy analysis. To assess corneal fluorescein clearing, the number and percentage of patients reaching a CFS score of 0 was calculated at each timepoint with the corresponding 95% confidence interval. Symptoms and TBUT were analyzed with a mixed model for repeated measures using the corresponding baseline value as covariate. Investigator global efficacy evaluation and Schirmer tear test without anesthesia were compared between groups using the Wilcoxon rank test.
Other analyses
Exploratory endpoints included changes in tear film osmolarity and impression cytology for HLA-DR expression using arbitrary units of fluorescence (AUF). The HLA-DR expression was summarized using descriptive statistics, and tear film osmolarity was compared between groups using the Wilcoxon rank test. In addition, mean change from baseline to day 84 in NEI-VFQ-25 subscale scores was compared between groups using an ANCOVA model with treatment as fixed effects and baseline value of the corresponding score as covariate. Safety variables were analyzed using descriptive methods.
Sample size
Sample size calculation was performed to test the noninferiority of CE vs HS on OSS. A standard deviation (SD) of 2.5 points was expected for this parameter. With an alpha risk of 0.025 (unilateral test of comparison vs the noninferiority margin), a power of 90%, and a noninferiority margin of 2 points, a sample size of 66 patients was found to be necessary. It was therefore planned to include 80 patients in this study, allowing for 15% of patients to be excluded from the PP set as a result of major protocol deviations.
Results
Patient demographics
A total of 85 patients were randomized to treatment and were included in the FAS: 44 patients in the CE group and 41 patients in the HS group. Most patients completed the study and only 5 patients withdrew from the study early. Eleven patients had major protocol deviations; therefore, 74 patients (37 in each group) were included in the PP set. The patient flow from screening until the completion of the study is summarized in Figure 1.
Patient flow during the NOSIKA study. AE = adverse event; CE = cationic emulsion; CI = confidence interval; FAS = full analysis set; FU = follow-up; GCP = Good Clinical Practices; HS = hyaluronate sodium; PP = per protocol.
Demographic and baseline characteristics were generally similar between the 2 treatment groups (Tab. I). Most patients were female (79.5% and 82.9% in the CE and HS groups, respectively) with a mean (± SD) age of 62.6 ± 13.2 years. Mean (± SD) duration of the disease for the analysis eye was slightly less in the CE group compared with the HS group (7.2 ± 5.6 years vs 9.9 ± 7.2 years, respectively).
Demographic and baseline disease characteristics (full analysis set)
BCDVA = best-corrected distance visual acuity; CE = cationic emulsion; DED = dry eye disease; HS = hyaluronate sodium; NEI VFQ-25 = National Eye Institute Visual Function Questionnaire (25 items); OSS = ocular surface staining; SD = standard deviation; TBUT = tear break-up time.
Clinical signs and symptoms were similar between groups and were indicative of moderate to severe symptomatic DED (Tab. I).
All of the patients in both groups had previously been treated with some form of dry eye therapy (data not shown).
Efficacy results
Results concerning noninferiority hypotheses are preferentially presented for the PP population and are supported by results in the FAS. Superiority hypotheses (secondary endpoints) are presented for the FAS population and are supported by findings in the PP population.
Objective clinical signs
Ocular surface staining score
In the PP set, after 1 month of treatment, mean OSS score improved by a similar magnitude in patients treated with either CE or HS (-2.5 ± 1.3 and −1.9 ± 1.6, respectively; Tab. II and Fig. 2). Mean (95% confidence interval) treatment difference of CE compared with HS was −0.80 (-1.45, −0.15), and CE was thus demonstrated to be noninferior to HS (p<0.0001) at day 28. This finding was supported by analysis of the FAS (Tab. II and Fig. 2).
Change in ocular surface staining score at day 28 (full analysis and per protocol sets)
ANCOVA = analysis of covariance; CE = cationic emulsion; CFB = change from baseline; CI = confidence interval; FAS = full analysis set; HS = hyaluronate sodium; LS = least squares; NI = noninferiority; OSS = ocular surface staining; PP = per protocol.
Noninferiority of cationic emulsion (CE) compared with hyaluronate sodium (HS) in ocular surface staining score at day 28. CI = confidence interval; FAS = full analysis set; PP = per protocol.
Improvement in OSS for both CE and HS groups was noted as early as day 7 and was maintained after 3 months of treatment at day 84. In the PP set, at day 7 the mean change from baseline in the OSS score was −1.7 ± 1.5 in the CE group compared with −1.2 ± 1.3 in the HS group. At day 84, this change was −2.8 ± 1.5 vs −2.2 ± 1.8 in the CE group and HS group, respectively. The CE treatment response was thus noninferior to HS throughout the study. Similar findings in the FAS confirmed these results.
Treatment-induced improvement in staining was similar for all areas of the eye. In the PP set, maximal mean (± SD) change from baseline for lissamine green staining of the nasal and temporal interpalpebral regions after 3 months was −1.0 ± 0.7 and −0.9 ± 0.6 in the CE group, respectively, vs −0.7 ± 0.8 and −0.6 ± 0.8 in the HS group, respectively. Mean (± SD) change from baseline in CFS after 3 months was −0.9 ± 0.8 vs −0.8 ± 0.9 for CE and HS, respectively. Similar findings in the FAS confirmed these results.
Corneal fluorescein clearing
The proportion of patients with complete corneal fluorescein clearing gradually increased from baseline and was numerically greater, but did not reach statistical significance in the CE group compared with the HS group at each study visit (Tab. III). At day 84, the proportion of patients with complete clearing was 31.8% and 22.0% in the CE and HS groups, respectively (p = 0.3383). A similar trend was noted for the PP population, with statistical significance almost achieved at day 28 (35.1% vs 13.5%, respectively; p = 0.0561).
Change from baseline for objective clinical signs not related to ocular surface staining (full analysis set)
CE = cationic emulsion; CFB = change from baseline; HS = hyaluronate sodium; SD = standard deviation; TBUT = tear break-up time.
Wilcoxon test was done because normality hypothesis not respected.
Other clinical signs
Schirmer test and TBUT scores were similar and not significantly different between CE and HS groups over the course of the study (Tab. III). After 3 months of treatment, the mean (± SD) change from baseline in Schirmer test score was 1.9 ± 3.6 mm/5 min vs 1.8 ± 3.6 mm/5 min in the CE and HS groups, respectively. At 3 months, mean (± SD) change from baseline in TBUT was 1.3 ± 2.4 seconds vs 1.1 ± 2.6 seconds, respectively.
Tear film osmolarity was investigated as an exploratory efficacy endpoint. No noticeable change from baseline or significant between-group differences in mean tear film osmolarity were observed (Tab. III).
Impression cytology was performed at baseline only and showed no relationship between HLA-DR and patient sex, although greater median values of HLA-DR expression were associated with premenopausal women vs postmenopausal women (63867 AUF vs 32520 AUF, respectively).
Subjective efficacy evaluations
Symptomatology
Global symptoms score improved in both groups over the duration of the study; however, symptom scores improved to a greater extent and after a shorter duration of treatment with CE compared with HS (Fig. 3). By day 28, total symptom improvement was significantly greater in the CE group compared with the HS group, with mean ± SD (%) VAS improvement of −14.8 ± 17.3 (-38.0%) vs −7.6 ± 14.2 (-19.9%), respectively, and a difference between groups of −6.8 ± 3.4 (p = 0.0469). The improvement in symptoms was maintained after 3 months and was numerically greater in the CE group than in the HS group, although this difference was not statistically significant.
Change from baseline of total ocular discomfort score at day 7, day 28, and day 84 (full analysis set). CE = cationic emulsion; HS = hyaluronate sodium; SE = standard error; VAS = visual analogue scale.
For individual scores, all symptoms except pain showed a greater improvement in the CE group as compared with the HS group. After 1 month, differences in both itching and eye dryness reached statistical significance. Itching improved, as measured by mean ± SD (%) VAS improvement of −14.8 ± 21.2 (-39.1%) in the CE group vs −1.7 ± 19.7 (-5.6%) in the HS group, with a mean ± SD treatment difference of −10.9 ± 4.1 (p = 0.0100). Eye dryness also improved, as measured by mean ± SD (%) VAS improvement of −21.9 ± 28.3 (-46.1%) vs −8.4 ± 21.4 (-17.0%) for CE and HS groups, respectively, with a mean ± SD treatment difference of −15.3 ± 4.7 (p = 0.0016). Improvements in both itching and eye dryness symptoms were maintained in the CE group at day 84. However, only itching was significantly better at this timepoint, with mean ± SD treatment differences of −12.0 ± 4.8 (p = 0.0135) and −9.1 ± 4.9 (p = 0.0676) for itching and eye dryness, respectively (Fig. 4, A and B).
Mean visual analogue scale (VAS) symptom scores for itching (
Investigator global assessment
Investigator global assessment of efficacy was rated as satisfactory for most patients in both treatment groups throughout the study. The proportion of patients whose treatment was rated as satisfactory was higher (though only significant at day 84) in the CE group than in the HS group at day 7 (69.8% vs 60.0%, respectively; p = 0.362), day 28 (85.7% vs 78.0%, respectively; p = 0.4052), and day 84 (87.5% vs 65.0%, respectively; p = 0.0339).
Quality of life
Overall, only small changes in mean global and individual NEI VFQ-25 scores were noted from baseline to day 84 in both treatment groups. At day 84, mean (± SD) changes from baseline values in the CE and HS groups were just 1.6 ± 9.2 and −1.3 ± 7.5, respectively (p = 0.1117). However, the change from baseline to day 84 for the individual scores of both ocular pain and driving function were significantly better in the CE group compared with the HS group. Mean score change for ocular pain was 10.0 ± 15.3 vs −0.9 ± 19.1 (p = 0.013) and for driving function was 2.7 ± 6.9 vs 3.4 ± 15.6 (p = 0.013) in the CE and HS groups, respectively.
Safety results
An overview of AEs is presented in Table IV. A total of 19 patients had a treatment-emergent AE during the study: 8 (18.2%) in the CE group and 11 (26.8%) in the HS group. Most AEs were mild or moderate and ocular in nature. All AEs deemed treatment-related or possibly treatment-related were ocular and were reported for 3 patients (6.8%) in the CE group and 4 patients (9.8%) in the HS group. The most common treatment-related or possibly treatment-related AE was instillation site pain, reported in 2 patients (4.5%) in the CE group and 1 patient (2.4%) in the HS group.
Overview of adverse events (full analysis set)
SAE = serious adverse event; TEAE = treatment-emergent adverse event.
Four patients had AEs that led to withdrawal: 3 (6.8%) in the CE group and 1 (2.4%) in the HS group. All of these AEs except one (influenza), in the CE group, were ocular in nature and related to study treatment. Related ocular AEs that led to withdrawal were 2 AEs of instillation site pain and 1 AE of instillation site erythema in the CE group and 1 AE each of conjunctival hyperemia and eye pain in the HS group.
One patient in each treatment group experienced 1 serious AE: these were influenza and femoral neck fracture, both of which were systemic and not related to study treatment. The patient with influenza (in the CE group) died as a result of this event.
No major differences between treatments were found with respect to local ocular tolerance parameters assessed during examination with the slit-lamp, and abnormalities were rarely observed. There was no change in best-corrected distance visual acuity or intraocular pressure in either treatment group over the course of the study (data not shown).
Discussion
The aim of this study was to compare the efficacy and safety of a CE with a well-established lubricant in patients with moderate to severe DED. The excellent track record of safety and efficacy of 0.18% HS solution in clinical studies (32-34), with its demonstrated superiority over saline solution in reduction of symptom frequency in patients with DED, made it a relevant choice as a reference comparator tear substitute in this study.
The results of this study showed that the preservative-free CE eyedrop caused a significant improvement of symptoms of DED as compared to HS; however, this difference was not observed in effect on the objective signs, and CE was therefore found to be noninferior to HS in reducing objective signs of keratitis or keratoconjunctivitis in DED.
The CE-induced improvement in OSS score was noticeable by day 7 and was maintained over 3 months of treatment; this effect was particularly apparent on the cornea, with almost one-third of patients showing complete corneal fluorescein clearing as soon as 1 month of treatment, although this difference was not statistically significant between HS and CE.
In addition to improved OSS scores, patients in the CE group reported a significantly greater reduction in global symptoms score of ocular discomfort compared with the HS group after 1 month of treatment, with the treatment effect in favor of CE still present after 3 months, although not statistically significant at this timepoint.
This study is the first to be conducted with CE in patients with moderate to severe DED. The results indicate that this CE has a beneficial effect on the ocular surface and are in agreement with other clinical studies conducted with this agent (27, 28).
In the present study, the improvements from baseline in TBUT and Schirmer test scores were similar in magnitude to those seen in a previous clinical study of CE (27). In this previous study, CE was compared with an aqueous solution containing polyvinyl alcohol and povidone in patients with mild to moderate DED; after 1 month of treatment, patients treated with CE demonstrated significantly greater improvements in global and individual symptoms scores and clinical signs (i.e., TBUT and lissamine green staining) compared with patients treated with the comparator (27).
In a second study assessing the efficacy of CE compared with an aqueous solution containing sodium CMC and glycerol and a nonionic emulsion formulation (EM) in patients with moderate DED over 3 months, both CE and CMC showed significant improvements in symptoms after 1 month, and all treatment groups had significantly improved symptoms by 3 months. Significant improvements were also noted in both TBUT and fluorescein staining for CE and CMC groups (but not the EM group) at 3 months. Notably, only CE significantly reduced tear film osmolarity compared with EM (28); however, no change in tear film osmolarity was noted with CE in the present study. It should be noted that in the present study, baseline values of tear film osmolarity were normal and could explain the absence of reduction of this parameter.
There are some limitations to this study. The single-masked design of this study is one such limitation; owing to the milky appearance of CE vs the clear HS solution, it would have been very difficult to use a double-masked design. However, investigators were masked to the study treatments to avoid any bias in the assessment of clinical parameters. Moreover, the relatively small sample size may explain some results observed; some evaluations (for example, the proportion of patients with complete corneal fluorescein clearing) appeared to show a clear numerical improvement in the CE group compared with the HS group, which did not achieve statistical significance with this sample size. Finally, similar to many clinical studies of DED, treatment response to both objective signs and subjective symptoms did not completely correlate (8, 35, 36). For example, significant improvements in symptomatology were noted for CE compared with HS, but objective test scores did not show a significant difference. Aside from the relatively small sample size as discussed earlier, this could be explained by the heterogeneity of the population enrolled in clinical studies, the variability of diagnostic tests used, or due to the short follow-up period (37, 38).
Despite these limitations, which are commonly encountered when designing and performing clinical trials in DED (37, 39), this study demonstrated the noninferiority of CE compared with HS on the improvement of OSS scores. However, CE showed greater improvement on global DED symptoms compared with HS, leading to benefits in patient quality of life.
The safety profiles of CE and HS were similar. The number of patients with at least 1 treatment-related ocular AE was low in both treatment groups and indicated that treatment with the CE was well-tolerated during the 84 days of study treatment.
In conclusion, this study has demonstrated that a CE, Cationorm®, a lipid-containing eye drop, was well-tolerated and effective in the treatment of patients with moderate to severe DED with keratitis or keratoconjunctivitis, showing that it is a valid option in this indication and may provide patients with a new therapeutic alternative.
Footnotes
Acknowledgment
The authors thank Scinopsis Medical Writing for their help with this manuscript; Chameleon Communications International for medical copyediting support; Maëva Deniaud for statistical advice; and Françoise Brignole-Baudouin and Luisa Riancho for providing cytometric analyses of impression cytology.
Financial support: The NOSIKA study was sponsored by Santen SAS, Evry, France.
Conflict of interest: M. Amrane, D. Ismail, and J.-S. Garrigue are employees of Santen. P.-Y. Robert is a clinical study investigator for Santen, Thea, Boiron, and Allergan. B. Cochener is a clinical study investigator and consultant for Alcon, AMO, RVO, Santen, Thea, and Physiol. P.-J. Pisella has no conflicts of interest. C. Baudouin is a consultant for, or has received research grants from, Allergan, Santen, and Thea, and is a consultant for Alcon.
Meeting presentation: Data from the NOSIKA study were presented at the International Symposium on Ocular Pharmacology and Therapeutics (ISOPT) Congress on June 19-22, 2014, in Reykjavik, Iceland.