Safety and Efficacy of Topical Infliximab in a Mouse Model of Ocular Surface Scarring

Abstract

Purpose.: To evaluate the safety/efficacy of topical infliximab, an anti-TNF-α monoclonal antibody, in a mouse model of ocular surface scarring.

Methods.: Twenty alkali burn mice were treated with infliximab (10 mg/mL) topically 6 times a day, while 20 alkali burn mice received saline for 7 days. Corneal opacity, epithelial wound healing, and ocular phimosis were examined at the slit-lamp. Tear production was quantified with phenol red thread test. Immunofluorescence for infliximab penetration, TNF-α localization, CD45+ cell infiltration, PAS, and Masson's trichrome staining were evaluated on ocular globes and eyelids. TNF-α and IL-1β expression levels were measured on treated murine corneas and eyelids. Finally, quantification of corneal CD31+ blood vessels and LYVE1+ lymphatic vessels were evaluated on 10 additional alkali burn mice receiving either infliximab or saline, after 14 days.

Results.: Topical infliximab penetrated the cornea and the conjunctiva and was not toxic (negative fluorescein stain). Its molecular target, TNF-α, was detected in the cornea after injury. Infliximab significantly reduced corneal perforation, opacity index, phimosis, leukocyte infiltration, and fibrosis in the eyelids. It also significantly prevented goblet cell infiltration in epithelial cornea and loss in the conjunctiva (P < 0.05), improved tear secretion and epithelial healing (P < 0.05). Finally, it significantly reduced both corneal hem- (P < 0.05) and lymphangiogenesis (P < 0.01).

Conclusions.: Infliximab penetrates the cornea and is safe to the ocular surface in an animal model of ocular surface scarring. We suggest that topical application of infliximab may be a useful treatment in ocular caustications.

Introduction

Ocular burns represent up to 18% of eye injuries referred to emergency departments and approximately 4% of all occupational injuries. They embody a significant clinical problem affecting prevalently young subjects. ^1,2 More than 80% of these injuries are due to chemical agents. Unlike the majority of ocular thermal burns, chemical burns cause severe complications; they often require multiple corneal grafting procedures (up to one-third of all corneal grafts), leave approximately 30% of patients disabled, and 15% blind. ³  

Ocular surface scarring is a common finding after ocular caustication. ⁴ Scarring, together with inflammation and neovascularization, is responsible for many unfavorable outcomes in burned eyes, including ocular perforation, low success rate of corneal transplantation procedures, and finally, impaired vision. 

Tumor necrosis factor-alpha (TNF-α) is a fibrogenic cytokine able to stimulate fibroblast replication. ⁵ Increased levels of TNF-α have been found in murine burned ⁶ and human corneas, ⁷ and a correlation between conjunctival cicatrization and this cytokine has been found in patients with trachoma. ⁸ Moreover, earlier work suggests that the soluble receptor to TNF-α (sTNFR1) may protect corneal fibroblasts from TNF-α toxicity by acting as an sTNF-α scavenger. ⁹  

On the other hand, recent evidence suggests TNF-α plays a role in corneal hemangiogenesis—its administration increased neovascularization in an alkali burn model. ¹⁰ Inhibition of TNF-α also inhibits choroidal neovascularization. ¹⁰ In fact, it has been hypothesized that TNF-α may induce expression of angiogenic factors and favor recruitment of macrophages, which could stimulate angiogenesis. ¹¹  

Infliximab is a chimeric IgG1κ monoclonal antibody. It neutralizes the biological activity of TNF-α by binding with high affinity to the soluble and transmembrane forms of TNF-α and inhibiting receptor binding. It is commonly administered systemically to treat Crohn disease, rheumatoid arthritis, and other autoimmune diseases. Topical treatment with infliximab also significantly improved chronic skin ulcers, with no reported systemic or local side effect. ¹² However, topical treatment at the normal ocular surface with infliximab has not been reported yet. Indeed, in light of potential clinical applications, safety of topically applied infliximab would be the obvious first concern. 

The aim of this study is to test whether topical blockade of anti-TNF-α (infliximab) is toxic to the normal ocular surface and whether it could ameliorate inflammation and scarring in a mouse ocular alkali burn model. 

Materials and Methods

Animals

Female, 4- to 6-week-old, C57BL/6 mice (Charles River Laboratories, Calco, LC, Italy) were used in all experiments (92 total mice or 97 total eyes, because 10 normal eyes from five mice were used as real-time PCR calibrators). Animals were allowed to acclimatize to their environment for 1 week prior to experimentation. Each animal was deeply anesthetized with intraperitoneal injection of tribromoethanol (250 mg/kg) before all surgical procedures. Postoperatively, all animals received a single dose of carprofen (Rimadyl; Pfizer, New York, NY) at 5 mg/kg SQ. Carbon dioxide inhalation and subsequent cervical dislocation was applied to euthanatize the animals. All experimental protocols were approved by the Animal Care and Use Committee of the San Raffaele Scientific Institute, in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Alkali Burn and Treatment

A corneoconjunctival alkali burn was created in the left eye of each mouse by means of a paper disc (3-mm diameter) soaked in 1 N NaOH for 10 seconds, under slit-lamp examination. The ocular surface was then washed with 15 mL of normal saline. To increase reproducibility, a single investigator applied the burn to all animals. Animals were then randomized into two groups: group 1 received 10 μL of 10 mg/mL infliximab (Remicade; Janssen Biologics B.V., Leiden, Holland) topically 6 times a day for 7 days (20 mice): 10 of these were used for immunostaining and 10 for real-time PCR analysis. In group 2 (10 mice), infliximab was administered for 14 days to measure corneal neovascularization. Finally, a control group (30 animals: 20 mice for 7 days and 10 for 14 days) received 10 μL of topical saline. 

One subconjunctival injection of infliximab (25 μL) was performed as a positive penetration control in three normal and three alkali burn eyes (n = 6 mice), and compared with subconjunctival injection of saline (n = 3). In addition, topical infliximab was tested in three eyes after mechanical removal of the corneal epithelium. 

Topical infliximab toxicity was evaluated in normal mice without alkali burns (n = 6) compared with saline-treated mice (n = 6). 

All treatments started immediately after caustication. Corneal photographs were taken with a biomicroscope using a digital camera (EOS 30D; Canon, Tokyo, Japan) attached to the slit-lamp microscope (Photoslitlamp, model 40 SL-P; Zeiss, Oberkochen, Germany) on the seventh day. 

Clinical Endpoints

Clinical examination was performed to check for toxicity induction and gross pathologic changes. Corneas were examined under the slit-lamp microscope (Photoslitlamp, model 40 SL-P; Zeiss) on day 7 after the alkali burn in a blind experiment. Fluorescein staining was used to evaluate infliximab toxicity and corneal epithelial wound healing after infliximab treatment. A scoring system was used to evaluate corneal opacity as previously described. ¹³ Ocular phimosis was scored on a scale of 0 to 3, where 0 = eye opens completely; 1 = eye opens more than a half; 2 = eye opens less than a half; and 3 = eye is completely closed. Tear production was quantified with phenol red thread test, as previously described. ¹⁴  

Immunohistochemical Analysis of Infliximab, CD45+ Cells, Pas Staining, and Fibrosis

On day 7, five of 10 eyes and eyelids per group were frozen in OCT compound (Killik; Bio-Optica, Milan, Italy) and stored at −80°C until ready for sectioning. Cryosections (7 μm) were fixed with paraformaldehyde (PFA) 4% (Sigma-Aldrich, St. Louis, MO) for 15 minutes at room temperature and rinsed in PBS (Sigma-Aldrich) three times. To block nonspecific staining, slides were incubated in 2% BSA (Sigma-Aldrich) and 0.3% nonionic surfactant (Triton X-100; Sigma-Aldrich) in PBS for 1 hour at room temperature. Alexa Fluor-546 goat antihuman monoclonal antibody against IgG (H+L) 2 mg/mL (Molecular Probes; Invitrogen, Eugene, OR) and Alexa Fluor-488 rat antimouse CD45 0.5 mg/mL (BioLegend, San Diego, CA) was applied to the eye and eyelid sections, respectively, in a 1:500 dilution for 2 hours at room temperature. Sections were counterstained with DAPI (Vector Laboratories, Inc., Burlingame, CA), mounted and photographed using a digital camera (Leica DFC310FX; Leica Microsystems, Wetzlar, Germany), attached to a fluorescence microscope (Leica CTR5500; Leica Microsystems). 

The remaining five eyes and eyelids per group were included in paraffin to detect goblet cells and fibrosis: paraffin sections were processed with periodic acid-Schiff's reagent (PAS) and Masson's trichrome staining, and counterstained with hematoxylin. Digital quantification of fibrosis was conducted as previously reported. ¹⁵  

Immunostaining of Corneal Neovascularization and TNF-α

On day 14, ten corneas per group (10 treated with infliximab and 10 with saline, total 20 corneas) were dissected and rinsed in PBS. The corneal epithelium was subsequently scraped off after EDTA (Sigma-Aldrich) treatment for 30 minutes at 37°C. Fixation of the tissue was conducted with iced acetone for 15 minutes following 2 hours of blocking in PBS/2% BSA. For visualization of blood and lymphatic vessels, corneas were immunostained with rat antimouse CD31 (1:200) 0.5 mg/mL (BioLegend) and goat antimouse LYVE-1 (1:200) 1 mg/mL (AbCam, Cambridge, UK) at 4°C overnight and subsequently with Alexa Fluor-594 donkey antirat IgG 2 mg/mL and Alexa Fluor-488 donkey antigoat IgG 2 mg/mL (Invitrogen) in a 1:500 dilution for 2 hours at room temperature. This was followed by three rinses in PBS. Corneas were flat-mounted using a mounting medium (VECTASHIELD; Vector Laboratories, Inc.) with DAPI and analyzed by fluorescence microscope (Leica CTR5500; Leica Microsystems). Digital pictures of the flat mounts were analyzed by a Java-based image processing program (ImageJ 1.44p; National Institutes of Health, Bethesda, MD). The total corneal area was outlined by using the innermost vessel of the limbal arcade as the border, and the area of neovascularization was then calculated and normalized to the total corneal area (expressed as a percentage of the cornea covered by CD31+ or LYVE1+ vessels). 

TNF-α localization was evaluated on corneal epithelium and stroma 7 days after alkali burn (n = 3 mice), as detailed above, with rat anti-mouse TNF-α (1:200) 0.5 mg/mL (BioLegend) and Alexa Fluor-488 donkey antirat IgG (1:500) 2 mg/mL (Invitrogen). The images were taken by confocal microscopy (model TCS SP5; Leica Microsystems). 

Analysis of Cytokine Transcripts by Real-Time PCR

On day 7, five pools of two mouse corneas or eyelids (upper and lower) per group were homogenized with a homogenizer (Ultra-Turrax T8; IKA, Wilmington, NC). Total RNA of mouse samples were extracted using TRIzol and quantified by spectrophotometry (NanoDrop ND 1000 Spectrophotometer; Thermo Scientific, Wilmington, DE). The ratio of absorbance values at 260 nm and 280 nm was used to assess purity of total RNA samples and a cutoff of >1.9 was applied. A DNAse treatment of RNA was performed with a DNA-free kit (Ambion, Austin, TX). RNA (1 μg) was transcribed into cDNA using a reverse transcription kit (MessageSensor RT; Ambion) and random hexamer primers (Applied Biosystems, Foster City, CA), according to the manufacturer's instructions. The PCR reaction was tested in a 7500 real-time PCR instrument using an amount of cDNA derived from 100 ng RNA; TaqMan Universal Master Mix II and Taqman Gene Expression Assays (Applied Biosystems) for TNF-α (Mm00443258_m1) and for interleukin-1β (IL-1β, Mm01336189_m1) were used. For assays, reactions were incubated at 95°C for 10 minutes, and then 45 cycles at 95°C for 15 seconds followed by 60°C for 1 minute. Each sample was analyzed as triplicates in a 20-μL volume. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH, Mm99999915_g1) transcript was used as endogenous control. A no-template control was included in each real-time PCR experiment to confirm the absence of DNA contamination. Results are presented as relative expression of 10 normal corneas (n = 5 mice) or eyelids pooled together (ΔΔC_T method). 

Statistical Analysis

χ² test (without Yates correction) was performed to compare corneal perforation susceptibility between infliximab-treated corneas and controls. Unpaired t-tests were used to evaluate the difference in corneal transparency and PAS staining in the cornea. One-way ANOVA, following Bonferroni post hoc tests, were performed to assess the difference in tear production, CD45+ cell infiltration PAS, and Masson's trichrome staining in conjunctiva. Two-way ANOVA, following Bonferroni post hoc tests, were used to analyze the rate of epithelial healing and the difference in ocular phimosis between groups. Statistical analysis of real-time PCR data was performed using unpaired t-tests. Significance was defined as a P value < 0.05. Results are presented as mean ± standard error of the mean (SEM). All data were processed using graphics and statistics software (GraphPad Prism 5.0; GraphPad Software, Inc., San Diego, CA). 

Results

Corneal Penetration of Topical Infliximab in Mouse Alkali Burn Eye

Immunostaining did not detect infliximab in tissue sections of treated eyes with intact corneal epithelium (Fig. 1A). Following the alkali burn, infliximab was detected in the stroma, specifically in the center of the cornea and in the conjunctiva (Fig. 1B). Application of infliximab on corneas without epithelium showed a diffuse staining in the outer stromal layers after 24 hours (Fig. 1C). A strong signal in the whole corneal thickness was detected after a single dose of 250 μg of infliximab injected subconjunctivally in both control (Fig. 1D) and alkali burn eyes (Fig. 1E). None of the untreated burned corneas (negative controls) showed immunoreactivity for infliximab (Fig. 1F), demonstrating specificity of the anti-IgG antibody used in this study. 

Figure 1

 

Immunostaining for infliximab and TNF-α in normal and alkali burn eyes following different treatment modalities. (A) When corneal epithelium was in place, no penetration of topical infliximab was detected. (B) Destruction of epithelium by alkali burn or (C) its removal promoted topical infliximab penetration into the corneal stroma (red staining) and the conjunctiva (white arrow). (D, E) Subconjunctival injection of infliximab also induced extensive corneal staining after alkali burn, compared with saline (F). The white box marks where the magnified photos at the bottom were taken. (G, H) Immunostaining for TNF-α (green) showing positive stain both in corneal epithelium and stroma after alkali burn (confocal microscopy images).

Immunostaining for TNF-α was positive in the epithelium (Fig. 1G) and stroma (Fig. 1H) after alkali burns. 

Clinical Evaluation of Topical Infliximab Treatment

Topical infliximab (10 mg/mL) treatment was not toxic to the ocular surface as assessed with a slit-lamp exam (Figs. 2A, 2B). The absence of fluorescein staining confirmed the nontoxic effect of infliximab in a healthy cornea (Figs. 2G, 2H). Moreover, topical infliximab improved corneal transparency following alkali burns (Figs. 2C, 2D), and a visual reduction of corneal neovascularization was evident with a slit-lamp (Figs. 2E, 2F). In addition, topical administration of infliximab increased the rate of epithelial healing (Figs. 2I, 2J)—the difference between treatment and control groups was statistically significant (P < 0.05) at day 7 (Fig. 2K), showing potential to promote corneal epithelial wound closure. PAS staining of corneal section showed significantly less (65%) goblet cells in the epithelium (Figs. 2L, 2M). No goblet cells were found in control corneas. 

Figure 2

 

Topical infliximab treatment was not toxic to normal eyes; it improved corneal clarity, reduced fluorescein staining, and perforation after alkali burn. (A, B) Topical application of infliximab 10 mg/mL to normal eyes did not induce any gross pathological changes to the ocular surface. No epithelial damage was observed with vital fluorescein staining (G, H). (C, D) Slit-lamp examination of corneal transparence and perforation rate ([C], arrow: perforation); (E, F) corneal neovascularization and (I, J) epithelial defects. (K) Topical infliximab significantly increased the rate of epithelial healing after 7 days of treatment, in comparison with saline. (L, M) Topical infliximab significantly reduced the number of PAS-positive goblet cells in the corneal epithelium after alkali burn on day 7. This was calculated as numbers of PAS-positive cells in corneal epithelium per field, at ×400 magnification. The arrows in (L) indicate the PAS-positive goblet cells in the corneal epithelium. Histograms represent mean values ± SEM; *P < 0.05.

Topical infliximab treatment decreased the perforation rate by approximately 50% (from 57.14% to 26.32%), which is statistically significant (P = 0.0489; Fig. 3A). In addition, infliximab treatment was associated with increased transparency. The corneal opacity index improved from 3.40 ± 0.22 (mean ± SEM, in untreated eyes) to 2.41 ± 0.34 in eyes treated with infliximab, which is statistically significant (P = 0.0484; Fig. 3B). 

Figure 3

 

Topical infliximab improved various clinical outcomes in the alkali burn model. (A) Corneal perforation rate was significantly reduced by approximately 50% following infliximab treatment. (B) Corneal opacity index was significantly reduced in infliximab versus saline-treated eyes. (C) Tear production was significantly reduced by half after alkali burn. Tear secretion in infliximab-treated eyes was improved and not significantly different from unburned eyes. (D) Alkali burn–induced ocular phimosis decreased in a time-dependent fashion. Infliximab treatment improved phimosis significantly, starting from day 4 onwards. Histograms represent mean values ± SEM; *P < 0.05.

Tear secretion was significantly (P < 0.05) reduced in the saline control group (1.31 ± 0.21 mm), but not in the infliximab group (1.71 ± 0.29 mm), as compared with the unburned eyes (2.39 ± 0.12 mm), suggesting less severe dry eye in the infliximab group (Fig. 3C). Although tear secretion was higher in the infliximab than saline-treated eyes, this did not reach statistical significance. 

Ocular phimosis was evaluated after alkali burns between groups at different points in time (Fig. 3D). At day 0, no significant differences were found between saline-treated and infliximab-treated groups. Ocular phimosis index was reduced more rapidly by infliximab than saline (from 2.39 ± 0.18 to 0.68 ± 0.23), and this was significant from day 4 onwards (P < 0.05). 

Topical Infliximab Is Protective on Conjunctival Goblet Cells

Infliximab-treated eyes showed three times more goblet cells than vehicle-treated eyes (P < 0.05). This was measured with PAS staining (calculated as percentage of PAS-positive area to the total area of epithelium). Both untreated and treated alkali burn eyes had goblet cell reduction compared with healthy eyes on day 7 (both P < 0.001). The goblet cell area was 30.45 ± 4.73% in the healthy group, 4.50 ± 1.62% in the saline-treated group and 13.21 ± 2.62% in the infliximab-treated group (Figs. 4A, 4B). 

Figure 4

 

Topical infliximab reduced goblet cell loss and CD45+ cell infiltration in conjunctiva. (A, B) PAS staining shows a significant reduction of goblet cells after alkali burn in the conjunctival epithelium. After topical infliximab treatment, the goblet cell area was three times higher than in saline-treated eyes, although it did not reach the level of the unburned control group. (C, D) Infiltration of CD45+ cells in the conjunctiva was significantly reduced in the infliximab-treated group on day 7. White arrows indicate positive staining for CD45+ cells. (E, F) Infliximab treatment significantly reduced fibrosis at the ocular fornix in comparison to saline on day 7. The fibrotic area was evaluated on Masson's trichrome stained cross-sections and quantified as the blue area at the fornix on both ocular side (between conjunctiva and sclera) and palpebral side (between conjunctiva and tarsus). Histograms represent mean values ± SEM; *P < 0.05, ***P < 0.001.

Inflammatory CD45+ Cell Infiltration Is Reduced in the Eyelid

The eyelids of alkali burn eyes showed an increase in inflammatory cell infiltration (positive staining for CD45 marker on cross-section) as opposed to healthy eyelids on day 7 (Figs. 4C, 4D). The average of CD45+ inflammatory cells increased five times following the alkali burned (9.19 ± 1.36) versus unburned eye (1.92 ± 0.24). This difference is statistically significant (P < 0.001). After infliximab treatment, the leukocyte infiltration decreased significantly (more than half) to 3.89 ± 0.81 CD45+ cells/field in comparison with saline-treated eyelids (P < 0.001). 

Alkali Burn Induces Gross Histological Changes in the Conjunctiva

Masson's trichrome staining showed an increase of fibrous tissue at the ocular fornix after injury (Fig. 4E). Infliximab treatment resulted in significantly reduced fibrosis (15% reduction, P < 0.05; Fig. 4F). 

Topical Infliximab Regulates Inflammatory Cytokine Expression in the Eyelid and Cornea

Both IL-1β and TNF-α were upregulated more than 1 log₁₀ after alkali burn in the eyelid; IL-1β was expressed more than 2 log₁₀ in the cornea, with respect to healthy eyes. IL-1β and TNF-α were reduced—by 72.3% and 56.5%, respectively—following infliximab treatment in eyelids (Figs. 5A, 5B). The transcript levels of IL-1β were decreased by 36.5% after infliximab treatment also in the cornea (Fig. 5C), although this reduction was not statistically significant. TNF-α transcript levels in corneas were not affected by either injury or treatment (Fig. 5D). 

Figure 5

 

Infliximab downregulated mRNA expression of IL-1β both in the eyelid and cornea, and TNF-α in the eyelid. (A, B) Transcript levels of IL-1β and TNF-α were increased in the eyelids of alkali burn mice as opposed to healthy mice, set at 1 (red line). After infliximab treatment, the expression levels of IL-1β and TNF-α were downregulated with respect to the saline-treated group. (C) The same result was obtained for IL-1β transcript in cornea. (D) TNF-α mRNA levels were not altered in the cornea after alkali burn and infliximab treatment. Histograms represent mean values ± SEM.

Topical Infliximab Effect on Corneal on Hem- and Lymphangiogenesis

Immunohistochemical staining of CD31 and LYVE1 on whole-mounted alkali burned corneas showed the growth of both blood and lymphatic neovessels in the alkali burn model. Topical treatment with 10 mg/mL infliximab six times daily for 14 days reduced the hematic- and lymphatic-neovascular areas when compared with saline (Figs. 6A, 6C). The treatment group (n = 10) showed a significant decrease in hemangiogenesis (P < 0.05) and lymphangiogenesis (P < 0.01) compared with the control group (Figs. 6B, 6D). 

Figure 6

 

Topical application of infliximab significantly decreased lymphatic and blood vessels in the cornea after 14 days of treatment. (A) CD31-stained (red) blood vessels and (C) LYVE-1–stained (green) lymphatic vessels. Neovascularization (NV) index is the area covered by blood/lymphatic vessels (%) normalized to the total corneal area. Quantification of NV index for (B) blood and (D) lymphatic vessels in an alkali burn-induced neovascularization assay. Histograms represent mean values ± SEM; *P < 0.05, **P < 0.01.

Discussion

Ocular burns represent a clinical challenge and are commonly associated with sight threatening complications such as corneal perforation, fibrosis, neovascularization, and symblepharon. Current medical treatment aims to control inflammation by judicious use of corticosteroids and, sometimes, prophylactic antibiotics. Corneal neovascularization and scarring are common complications of ocular burns and frequently require corneal grafting to improve vision. Unfortunately, the outcomes of graft procedures on vascularized beds are poor, with a rejection rate above 50%. ¹⁶  

Topical application of infliximab would first require evidence of its safety and penetration through the normal cornea. For this reason, we tested whether infliximab could penetrate through the ocular surface barrier. Similar to what has been reported for bevacizumab, ²² we found that topical infliximab was not able to penetrate the ocular surface in normal eyes. This is not surprising, since infliximab has a similar molecular weight (149 KDa). However, following removal of the corneal epithelium or alkali burn, infliximab was found in the corneal stroma and in the conjunctiva, as occurs after subconjunctival injection. Interestingly, higher concentrations of infliximab were achieved in the cornea of burnt eyes as compared to normal eyes, regardless of the administration route (subconjunctival or topical). This suggests that inflammation-induced disruption of the ocular surface integrity favors antibody penetration. Although ocular penetration of antibody fragments have been described, ¹⁷ there has been no report in the literature regarding infliximab, which is currently approved for intravenous use. Finally, we did not detect any pathological changes—including fluorescein vital staining—following topical application of infliximab in normal eyes, suggesting that infliximab is safe. Moreover, we showed that topical infliximab is effective in reducing the main unfavorable sequelae of acute inflammation such as fibrosis, neovascularization, and corneal perforation, when compared with saline treatment. 

Recent work has shown that TNF-α inhibition can reduce corneal hemangiogenesis in an alkali burn mouse model. Differently from these authors, who used etanercept—a fusion protein which binds both TNF-α and TNF-β—we used infliximab, a monoclonal antibody which specifically binds TNF-α. Moreover, we checked both hemangiogenesis and lymphangiogenesis and found that lymphangiogenesis is dramatically affected by TNF-α inhibition (more than 50% reduction). We suggest that this finding may have significant clinical implications, when considering that alkali burn patients oftentimes undergo repeated corneal grafting. Indeed, it is a well-known fact that corneal neovascularization favors corneal graft rejection. ^16,18 Interestingly, it has been suggested that lymphatic neovessels may represent a serious threat for the induction of graft rejection. ¹⁹  

TNF-α plays a key role in the promotion of fibrosis; however, its precise mechanism of action has not been fully elucidated yet. It has been proposed that it has a profibrotic and proinflammatory effect on conjunctival fibroblasts, ²⁰ and also induces contraction of conjunctival fibroblasts. ²¹ Moreover, TNF-α induces the expression of the proinflammatory cytokine macrophage migration inhibitory factor by conjunctival fibroblasts. In this vein, we observed that ocular fibrosis at the fornix was reduced following infliximab treatment and that this was associated with a reduction of inflammatory cell infiltration in the conjunctiva. 

This study shows that topical blockade of TNF-α with infliximab is able to reduce inflammation and its sequelae in the cornea, conjunctiva and eyelids. This was observed clinically in-vivo, with increased corneal transparency, and ex vivo with reduced inflammatory CD45+ cell infiltration in the eyelids, and the expression of inflammatory cytokines in both the corneas and eyelids. Interestingly, although TNF-α increased in the eyelids after the alkali burn, we did not find an increase of TNF-α mRNA expression in the cornea, in line with previous reports. ²² However, the presence of TNF-α was detected in the cornea after the alkali burn, in line with previous reports. ¹¹ This confirmed the presence of a target for the topically administered infliximab. 

Conjunctiva and eyelids also appeared favorably affected by infliximab treatment. In fact, we observed reduced phimosis and conjunctival fibrosis in infliximab-treated eyes. This suggests that the reduced incidence of corneal perforations and increased corneal transparency may be a consequence of TNF-α blockade in the pericorneal milieu, such as the conjunctiva and eyelids. Indeed, alterations of the fornix and eyelid anatomy are common in ocular scarring disease and are frequently associated with corneal complications, such as punctate keratitis, corneal ulcers, and scarring. ²³  

Dry eye is a well known and feared complication in ocular cicatricial diseases, ²⁴ as it significantly reduces the quality of life. Topical infliximab was protective on conjunctival goblet cells and resulted in increased tear secretion, suggesting a potential role for this treatment in dry eye associated with ocular surface scarring. 

In summary, topical treatment with infliximab is safe and can penetrate the inflamed ocular surface. It is also effective in reducing ocular surface scarring sequelae on the whole ocular surface unit, including cornea, conjunctiva, and eyelids. 

Chemical burns of the cornea are among the most severe ocular injuries and treatment options are currently limited. We suggest that topical application of infliximab could represent a useful, adjunctive tool in the treatment of ocular surface inflammation and scarring. We cannot exclude that systemic absorption occurred; however, this is generally lower than with systemic administration. ²⁵ Further studies are underway to translate these findings to the clinic. 

Footnotes

 Supported by a grant from the Bietti Eye Foundation, Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS), Rome, Italy (GF).

Footnotes

 Disclosure: G. Ferrari, None; F. Bignami, None; C. Giacomini, None; S. Franchini, None; P. Rama, None