Effect of Topical Losartan in the Treatment of Established Corneal Fibrosis in Rabbits

Abstract

Purpose: The purpose of this study was to evaluate the safety and efficacy of topical losartan in the therapeutic treatment of established corneal scaring fibrosis at 1 month after alkali burn in rabbits.

Methods: Standardized alkali burns were performed in 1 eye of 24 rabbits with 0.75N NaOH for 15 seconds. Corneas were allowed to heal and develop scaring of the cornea for 1 month. Twelve eyes per group were treated with 50 µL of topical 0.8 mg/mL losartan in balanced salt solution (BSS), pH 7.0, and 12 eyes were treated with vehicle BSS 6 times per day. Six corneas were analyzed at 1 week or 1 month in each group. Standardized slit lamp photographs were obtained at the end point for each cornea and opacity was quantitated using ImageJ. Corneoscleral rims were cryofixed in optimum cutting temperature (OCT) solution and combined duplex immunohistochemistry for myofibroblast marker alpha-smooth muscle actin (α-SMA), mesenchymal cell marker vimentin, and TUNEL assay for apoptosis was performed on all corneas.

Results: Topical losartan was effective in the treatment of established stromal fibrosis following alkali burn injury to the rabbit cornea. Stromal myofibroblast density was decreased and stromal cell apoptosis was increased (included both α-SMA-positive myofibroblasts and α-SMA-negative, vimentin-positive cells) at both 1 week and 1 month in the topical losartan-treated compared with vehicle-treated groups.

Conclusions: Topical losartan is effective in the treatment of established stromal fibrosis in rabbits. Most myofibroblasts disappear from the stroma within the first month of losartan treatment. Longer treatment with topical losartan is needed to allow time for corneal fibroblast regeneration of the epithelial basement membrane (in coordination with epithelial cells) and the removal of disordered extracellular matrix produced by myofibroblasts.

Introduction

Several recent case reports in humans^1–3 and studies in rabbits^4–8 have noted the efficacy and safety of treatments with topical losartan to prevent or treat corneal scaring fibrosis. Rabbit studies performed thus far have used the topical losartan at the time of the injury to inhibit the development of myofibroblasts and opacity,^4–8 whereas in each of the human case reports, the scar was already present at the time the topical losartan treatment was initiated. 

The purpose of the present rabbit study was to trigger the development of myofibroblasts and fibrosis with alkali burns for 1 month after the injury and then to study the effects of the topical losartan treatment on established myofibroblasts and fibrosis. The stromal apoptosis response to topical losartan treatment compared to the vehicle treatment was also studied. 

Methods

All rabbit procedures and treatments were approved by the Institutional Animal Care and Use Committee at the Cleveland Clinic Foundation and the Animal Care and Use Review Office (ACURO) of the Department of Defense. All animals were treated in accordance with the tenets of the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Twenty-four female 12- to 15-week-old New Zealand white rabbits weighing 2.5 to 3 kg each were included in this study. Male rabbits were not included due to frequent fighting and resulting eye injuries noted in prior studies. 

All rabbits received 60 mL children's liquid acetaminophen (Johnson & Johnson, Ft. Washington, PA, USA) in each liter of drinking water 24 hours prior to alkali burns and continuing at least 5 days after the alkali burn. One eye of each rabbit was randomly selected to have the alkali burn injury and the rabbit was placed under general anesthesia with 30 mg/kg ketamine hydrochloride and 5 mg/kg xylazine by intramuscular (IM) injection. The eye to be treated received 2 drops of topical 1% proparacaine hydrochloride (Alcon, Fort Worth, TX, USA) prior to the injury. 

In the treated eye, an alkali burn was performed with 30 microliters of 0.75N NaOH solution in balanced salt solution (BSS) on a 5 mm diameter circle of Whatman #1 filter paper for 15 seconds, as previously described.⁹ This NaOH exposure reproducibly produces an immediate epithelial defect and injury to the underlying stroma, and a 1 to 2 mm central injury to the corneal endothelium, that increases the overall stromal wound healing response.⁹ It does not produce neovascularization that a 1N exposure produces that tends to be more variable between corneas.⁹ 

All eyes were treated with one drop of 0.3% ciprofloxacin (Alcon Laboratories, Ft. Worth, TX) 4 times a day until the epithelium closed or with ongoing treatment if a persistent epithelial defect (PED) developed. No mitomycin C or corticosteroids were administered. The rabbits were all allowed to heal for 1 month without topical losartan treatment but did receive ongoing topical ciprofloxacin drops if a PED was present. 

At 1 month after injury, rabbits were divided into 2 groups of 12 rabbits each receiving BSS vehicle, pH 7.0, or 0.8 mg/mL losartan (pure USP grade losartan powder available from PCCA (https://pccarx.com) in the United States) in BSS, pH 7.0, to the treated eye 6 times per day (approximately 8 AM, 10 AM, 12 noon, 2 PM, 4 PM, and 6 PM). Anesthesia was not used during losartan drop administration. 

Closure of Corneal Epithelial Defects

All corneas were monitored for epithelial defects with 1% fluorescein drops (Bausch & Lomb, Rochester, NY, USA) and a cobalt blue filter at weekly intervals. Findings were recorded and photographed at the end point for each cornea. 

Slit Lamp Photography and Opacity Analyses

All pupils were dilated with 2 drops of 1% tropicamide (Akorn Co., Lake Forest, IL, USA) for 30 minutes and slit-lamp photographs were obtained, without and with 1% fluorescein drops, with the rabbits under general anesthesia. Slit lamp photographs were taken at a standardized illumination level, angle of illumination and imaging angle, with 10 × magnification, using a Topcon (Oakland, NJ, USA) SL-D7 slit-lamp photography system. 

All slit lamp images for each cornea at the 2 time points were converted to 600-pixel width × 400-pixel height 300 dots per inch (DPI) images with Photoshop version 23.5.2 (Adobe, San Jose, CA) and then a 400 pixel × 400-pixel region was selected to generate the composite image figures and to prepare images for quantitation using ImageJ version 1.53a analysis software (https://imagej.net), as previously reported.¹⁰ Total corneal opacity in each group was determined from a 6 mm diameter central corneal circle in ImageJ placed just outside the light reflex of each cornea, with examples shown for each cornea in Figures 1 and 2. 

Figure 1.

 

Slit lamp images of each cornea 1 month after alkali burn prior (pre) to starting treatment with BSS vehicle or 0.8% losartan in BSS 6 times per day and after 1 week of treatment. The dotted yellow circle is the approximate area of ImageJ analysis of the opacity. One cornea in the losartan group (#4) developed a PED (arrows) during the interval between the alkali burn and beginning topical losartan treatment that persisted at the 1-week time point for euthanasia. It can be noted that the opacity tended to decrease in both the vehicle-treated and the losartan-treated corneas during the week of treatment. Magnification = 20×.

Figure 2.

 

Slit lamp images of each cornea 1 month after alkali burn prior (pre) to starting treatment with BSS vehicle or 0.8% losartan in BSS 6 times per day and after 1 month of treatment. The dotted yellow circle is the approximate area of ImageJ analysis of the opacity. One cornea in the losartan group (#1) developed a PED (arrows) during the interval between the alkali burn and beginning topical losartan treatment but the PED had resolved by the end of the month of treatment. A transient PED may have also developed in losartan cornea #6, but it was not present at the time of weekly screenings with fluorescein. It can be noted that the opacity tended to decrease in both the vehicle-treated and the losartan-treated corneas during the month of topical treatment. Magnification = 20×.

Corneal Cryofixation and Sectioning

At the end of the topical losartan treatment period (6 rabbits at 1 week and 6 rabbits at 1 month in each group), each rabbit was euthanized while under ketamine-xylazine general anesthesia with 100 mg/kg Beuthanasia (Shering-Plough, Kenilworth, NJ, USA) by intravenous injection, followed by exsanguination. Corneo-scleral rims were excised without touching the corneal surfaces with sharp Westcott scissors (Fairfield, CT, USA) and 0.12 forceps (Storz, St. Louis, MO, USA). Each corneo-scleral rim was centered in a 24-mm × 24-mm × 5-mm mold (Fisher Scientific, Pittsburgh, PA, USA) and the mold was then filled with liquid optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA, USA). The mold and cornea were immediately quick frozen on dry ice and stored at −80°C until cryo-sectioning of the frozen OCT-cornea blocks was performed. 

Blocks were bisected in the center of the original alkali burn zone of each cornea and 10-µm-thick transverse sections were cut from the central cornea using a cryostat (HM 505M; Micron GmbH, Walldorf, Germany). Three sections from each cornea were placed on each 25-mm × 75-mm × 1-mm Superfrost Plus microscope slide (Fisher Scientific) and slides with sections were maintained at −20°C until immunohistochemistry (IHC) was performed. 

Immunohistochemistry for α-SMA and Vimentin, and TUNEL Assay

Duplex IHC was performed for α-SMA (myofibroblast marker, red) and vimentin (mesenchymal cell marker, at a concentration where all corneal fibroblasts and myofibroblasts, but only rare keratocytes stain positive,¹⁰ yellow), and simultaneous TUNEL assay (S7110, green; Sigma-Aldrich, Burlington, MA, USA)¹⁰ according to the manufacturer's instructions. Primary antibodies had been confirmed by Western blotting and IHC to recognize rabbit antigens (see the Table). Isotypic nonspecific control antibodies (ThermoFisher Scientific, Waltham, MA, USA) and secondary fluorescent tagged antibodies were used (see the Table). The 4′,6-diamidino-2-phenylindole (DAPI) was included in each IHC. Images were obtained at 100 × total magnification on a Leica DM6B upright microscope equipped with an automated stage and Leica 7000 T camera using the LASX software (Leica Microsystems, GmbH, Wetzlar, Germany). 

Table.

 

Primary and Secondary Antibodies Used in Immunohistochemistry

 

All images for each cornea at the two time points were converted to 600-pixel width × 400-pixel height 300 DPI images with Photoshop version 23.5.2 (Adobe, San Jose, CA, USA) to generate the composite image figures and to prepare images for quantitation using ImageJ version 1.53a analysis software (https://imagej.net), as previously reported.¹⁰ 

Myofibroblast α-SMA staining intensity in the stroma was quantitated in 120,000 pixels² total volume of stroma with the anterior edge of the ImageJ quantitation box placed just posterior to the EBM (or where EBM would be located) and the posterior edge of the quantitation box placed just anterior to the corneal endothelium. The width of the ImageJ quantitation box was adjusted to include 120,000 pixels² total volume of stroma for each cornea and, therefore, varied with the thickness of each cornea as it was placed on the IHC slide. Similarly, individual cells that were TUNEL-positive were determined as total cells per 120,000 pixels² volume box placed in the same stromal locations as for α-SMA. For both α-SMA and TUNEL quantitation, the results from four independent runs were averaged to provide the result for each cornea. 

Statistics

Statistical analyses were performed using the nonparametric Mann Whitney U test (https://www.statskingdom.com/kruskal-wallis-calculator.html) and P < 0.05 was considered statistically significant. 

Results

Figures 1 and 2 show the standardized slit lamp results at 1 month after injury pretreatment and after treatment with vehicle or topical losartan for 1 week or 1 month, respectively. All corneas in both the vehicle and losartan groups tended to have a decrease in ImageJ quantitated corneal opacity over either 1 week or 1 month of treatment (see Figs. 1, 2). The changes in opacity quantitated with ImageJ over the treatment intervals for both groups are shown in Figure 3. The mean change in ImageJ opacity between the vehicle group and the losartan group at either 1 week (−12.8 ± 3.1 vs. −21.3 ± 4.1 density units, respectively) or 1 month (−24.7 ± 3.8 vs. −24.0 ± 3.7 density units, respectively), did not reach statistical significance (P = 0.23 at 1 week and P = 0.87 at 1 month). One cornea in the 1-week topical losartan group continued to have a PED after losartan treatment but analyses with this cornea excluded did not change the statistical outcomes with regard to corneal opacity changes over time. 

Figure 3.

 

ImageJ analysis of the changes in central corneal opacity in all eyes from 1 week or 1 month of treatment with either vehicle or losartan. Blue is post alkali burn opacity at 1 month and prior to topical treatment, and red is the final opacity just prior to euthanasia. Green dots are the values for one cornea with a PED. Note the overall trend toward lower opacity in the losartan treated corneas but these differences did not reach statistical significance over this time interval. As expected, the stromal opacity was greatest in one cornea in the 1-week losartan group that had developed a PED, and in this cornea also the opacity was higher pretreatment with losartan than after 1 week of treatment.

Figure 4 (after 1 week of treatment) and Figure 5 (after 1 month of treatment) show representative triplex IHC for corneas in each group at each time point. In both groups, at both time points, a large proportion of the myofibroblasts detected were localized to the posterior half of the cornea, with a few exception examples shown in individual panels in Figures 4 and 5. Cells that were TUNEL-positive in the stroma were primarily α-SMA-positive myofibroblasts or α-SMA-negative, vimentin-positive cells (that were likely corneal fibroblasts, fibrocytes, or their daughter cells) that had begun differentiation into myofibroblasts but did not yet express α-SMA. 

Figure 4.

 

Duplex immunohistochemistry combined with the TUNEL assay in the 1-week vehicle- or losartan-treated corneas. For each cornea, a composite is shown, in addition to IHC for myofibroblast marker α-SMA (arrowheads), TUNEL assay to detect apoptosis (arrowheads), and IHC for mesenchymal marker vimentin. The red dotted lined rectangles are examples of the approximate quantitation boxes for red α-SMA signal that were 120,000 pixels² in each cornea but the actual measurement was performed with the image in ImageJ so the area of the analysis could be controlled. The green dotted rectangles are representative 120,000 pixels² stromal quantitation boxes for green TUNEL-positive cells that were counted individually. Notice, in both groups, more of the myofibroblasts detected are in the posterior cornea, with only a few detected in the anterior half of the stroma in some corneas. Cells undergoing apoptosis were identified at the microscope as a combination of α-SMA-positive, vimentin-positive myofibroblasts and α-SMA-negative, vimentin-positive cells that were likely corneal fibroblast, fibrocytes, and the daughter cells of these precursors that have begun their transition into myofibroblasts. No epithelial cells undergoing apoptosis were counted because apoptosis is a normal part of epithelial maturation. The magnification bar shown is representative of all panels.

Figure 5.

 

Duplex immunohistochemistry with combined TUNEL assay in the 1-month vehicle- or losartan-treated corneas. For each cornea, a composite is shown, in addition to IHC for myofibroblast marker α-SMA (arrowheads), TUNEL assay to detect apoptosis (arrowheads), and IHC for the mesenchymal marker vimentin. The red dotted lined rectangles are representative quantitation boxes for red α-SMA signal that were 120,000 pixels² in each cornea but the actual measurement was performed with the image in ImageJ, so the area of the analysis could be controlled. The green dotted rectangles are examples of the 120,000 pixels² stromal quantitation boxes for green TUNEL-positive cells that were counted individually in each box. Notice, in both groups, more of the myofibroblasts detected are in the posterior cornea, with only a few detected in the anterior half of the stroma in some corneas. Cells undergoing apoptosis were a combination of α-SMA-positive, vimentin-positive myofibroblasts and α-SMA-negative, vimentin-positive cells that were likely corneal fibroblast, fibrocytes, and the daughter cells of these precursors that have begun their transition into myofibroblasts. No epithelial cells undergoing apoptosis were counted because apoptosis is a normal part of epithelial maturation. The magnification bar shown is representative of all panels.

Figure 6 shows the α-SMA (level of stromal staining) and TUNEL (number of TUNEL-positive cells) in the analyzed 120,000 pixels² volume of stroma analyzed with ImageJ. Note that stromal myofibroblast α-SMA was significantly greater in the vehicle-treated compared to the losartan-treated group and the TUNEL-positive stromal cells were greater in numbers in the losartan-treated corneas than the vehicle-treated corneas at both time points. 

Figure 6.

 

Myofibroblast and other stromal cell apoptosis after 1 week or 1 month of treatment with vehicle or losartan. (A and B) The stromal myofibroblast signal was significantly greater in the vehicle-treated groups than the losartan-treated groups after both 1 week and 1 month of treatment. (C and D) TUNEL-positive stromal cells were greater in the losartan-treated groups than the vehicle-treated groups at both the 1 week and 1 month of treatment time points.

Discussion

This study in rabbits confirms the hypothesis that topical losartan can be used therapeutically to treat corneal scarring fibrosis that is already established after an injury, similar to each of the case reports published thus far for humans,^1–3 in which the topical losartan was used months to years after the original scarring injury. In each of the previously published rabbit studies,^4–8 the topical losartan treatment was begun at the time of injury and, therefore, evaluated prophylactic treatments for corneal scarring. The longest period after injury thus far reported for successful use of topical losartan to treat corneal stromal fibrosis in humans was 3 years after herpes zoster-induced scarring (R. Ambrosio, Jr., personal communication, 2023). It is our hypothesis, that myofibroblast-generated scars may respond to topical losartan even many years after the inciting injury because fibrosis in the stroma is actively maintained by persistent myofibroblasts that obtain requisite TGF-beta signaling for survival and ongoing production of disordered ECM from the tears and epithelium through a persistently dysfunctional EBM.^2,10–12 Once the myofibroblasts, and possibly their precursor cells have undergone losartan-induced apoptosis, corneal fibroblasts can occupy the fibrotic tissue to both (1) reabsorb and reorganize the disordered collagens and other matrix materials that contribute to the opacity in the fibrosis, but also to (2) coordinate with the overlying epithelium to regenerate the EBM and its normal TGF-beta regulatory function, so topical losartan treatment can be discontinued without recurrence of the scarring fibrosis.^12,13 

Despite the marked decrease in the number of myofibroblasts and higher level of stromal cell apoptosis in corneas treated with topical losartan compared with vehicle (see Figs. 4 to 6), at both the 1 week and 1 month of treatment time points, the decrease in opacity after 1 month of topical losartan treatment was small, indicating most of the opacity was due to remaining disordered ECM that would be removed more slowly over 6 or more months during the ongoing losartan treatment. This provides further support for the hypothesized “two-phase mechanism” of topical losartan fibrosis treatment¹⁴ in which the losartan, in phase I, triggers myofibroblast, and possibly myofibroblast precursor apoptosis directly by inhibiting signaling molecule ERK activation and, in phase II, inhibits further myofibroblast differentiation, whereas corneal fibroblasts facilitate regeneration of the EBM and remove and reorganize the disordered collagens and other ECM that were deposited by the myofibroblasts.^11,12 This phase II tends to be much longer and more variable between the eyes of different individuals with identical injuries and emphasizes the need for longer topical losartan treatment after clearing of the scarring. The fibrosis will tend to recur after discontinuation of the medication if the EBM has not fully regenerated. In some corneas, full regeneration of the EBM may never occur for unknown reasons, and myofibroblasts and opacity will tend to recur in these corneas once topical losartan treatment is discontinued despite many months of therapy. These corneas will require surgical treatments, such as phototherapeutic keratectomy or corneal transplantation. 

Even though the alkali burn used in this study⁹ has been shown to injure both the anterior and posterior stroma, the majority of myofibroblasts detected after 1 week or 1 month of treatment were in the posterior cornea in both losartan- and vehicle-treated groups. The likely interpretation of what was observed was that the EBM fully regenerated in most of these corneas during the month after alkali injury before treatment was started and/or during the topical treatment period that followed. Therefore, most myofibroblasts generated by the injury to the anterior stroma had already undergone spontaneous apoptosis when deprived of TGF beta by the regenerated EBM. Descemet's membrane, however, had not had time to regenerate and, therefore, TGF beta-1 and TGF beta-2 continued to enter the posterior stroma from the aqueous humor and maintain myofibroblasts in that area.^15,16 

In summary, this study shows that topical losartan provides safe and effective treatment for established corneal stromal fibrosis in a rabbit model. It also provides support for the “two-phase mechanism” of topical losartan action in which, phase I myofibroblast and myofibroblast precursor apoptosis is triggered by losartan inhibition of signaling molecule ERK, and phase II corneal fibroblasts slowly reabsorb and reorganized the disordered ECM while simultaneously cooperating with the corneal epithelium to regenerate the EBM. 

Acknowledgments

Supported in part by Department of Defense grant VR210001, US Public Health Service grant P30 EY025585 from the National Eye Institute, National Institutes of Health, Bethesda, Maryland, and Research to Prevent Blindness, New York, New York, and The Cleveland Eye Bank Foundation, Cleveland, Ohio. 

Proprietary Interest Statement: Steven E. Wilson and the Cleveland Clinic have submitted patents on the use of topical losartan and other angiotensin II receptor blockers to prevent and treat corneal scarring fibrosis. None of the other authors have any commercial or proprietary interest in this study. 

Disclosure: V.V. Martinez, None; B.A.L. Dutra, None; M.R. Santhiago, None; S.E. Wilson, Wilson and Cleveland Clinic are patent owners and licensers of topical losartan