Successful Regression of Newly Formed Corneal Neovascularization by Subconjunctival Injection of Bevacizumab in Patients with Chemical Burns

DOI: https://doi.org/10.21203/rs.3.rs-1852977/v1

Abstract

Purpose: To investigate the effect and timing of subconjunctival bevacizumab injection on inhibiting corneal neovascularization (CNV) in patients after chemical burns.

Methods: Patients with CNV secondary to chemical burns were involved. Two subconjunctival injections of bevacizumab (2.5 mg/0.1 mL per involved quadrant) with an interval of four weeks were administered, and followed up a year. The area occupied by neovascular vessels (NA), accumulative neovascular length (NL), mean neovascular diameter (ND), best-corrected visual acuity (BCVA) and intraocular pressure (IOP) were evaluated.

Results: Eleven patients with CNV were involved. Eight patients had a history of surgery (four had amniotic grafts, one had keratoplasty, and three had amniotic grafts and keratoplasty). Decreasing in NA, NL, and ND were statistically significant at each time point compared to the baseline (p < 0.01). CNV that developed within one month was considerably regressed, and vessels with fibrovascular membranes were found to be narrower and shorter than pretreatment. BCVA improved in five patients (from one to five lines), remained unchanged in five patients, and decreased in one patient compared to pretreatment.

Conclusion: Subconjunctival bevacizumab injection has a particular potential for the regression of CNV, especially within one month in patients after chemical burns.

Introduction

Corneal neovascularization (CNV) can lead to limbal stem cell deficiency, significant visual impairment, and subsequent high-risk keratoplasty with poor prognosis.1 The current methods for treating CNV include the use of steroids,2 laser photocoagulation,3 fine needle diathermy,4 and photodynamic therapy.5 The effects of these treatments are variable in humans.

Recently, neutralization of vascular endothelial growth factor (VEGF) has been demonstrated to effectively inhibit angiogenesis and promote corneal graft survival in animal models.6 In addition, topical and subconjunctival administration of the anti-VEGF agent bevacizumab (Avastin), as an off-label treatment, has been used to inhibit CNV in patients with herpes simplex keratitis,7 lipid keratopathy,8 corneal transplantation,9 bullous keratopathy,9 Stevens–Johnson syndrome,10 and pterygium.11 In another study, subconjunctival administration of bevacizumab was found to effectively inhibit CNV in the rabbit corneal alkali burn model.12 It was also shown that aflibercept was superior to bevacizumab in a rat chemical burn-induced neovascularization model;13 however, there was no CNV regression after treatment with aflibercept, betamethasone, or their combination14 in the rabbit corneal alkali burn model. Results generated in animal models have been inconsistent, and clinical observations have revealed that neovascular vessels do not disappear in most patients receiving bevacizumab treatment. A previous study has suggested that topical bevacizumab treatment can improve the overall graft success rate by preventing neovascularization.15

Some authors have suggested that early application of anti-VEGF agents may be a better approach for preventing CNV. However, no study has revealed a definitive intervention time after chemical burns. Therefore, this study aimed to investigate the efficacy and safety of subconjunctival bevacizumab treatment in patients with chemical burn-induced CNV, and to determine the appropriate timing for its administration.

Patients And Treatments

Patients 

This uncontrolled interventional study was conducted in the Corneal Division of the Zhongshan Ophthalmic Center at Sun Yat-sen University between March 2011 and July 2018. The study protocol conformed to the tenets of the Declaration of Helsinki and was approved by the medical ethics committee of the Affiliated First Hospital of Guangdong Pharmaceutical University (202263). Informed consent was obtained from all patients before the study began. 

The inclusion criteria for enrollment were as follows: (1) a history of ocular surface burns, (2) superficial or stromal CNV extending > 2 mm from the limbus, and (3) surgical or no surgical intervention. The exclusion criteria were as follows: (1) current or recent (≤3 months) corneal and/or ocular surface infection (with bacteria, viruses, fungi, or Acanthamoeba); (2) a history of any other treatment with anti-VEGF agents (topical or systemic); and (3) pregnancy, uncontrolled hypertension, and a history of thromboembolic events. 

Eleven consecutive patients (nine men and two women) who met the above criteria were enrolled in this study. The age range of the patients was 18–48 years (mean age = 32.64 ± 10.30 years). Among these patients, three had not undergone any surgery before treatment, one had a history of corneal transplantation, four had a history of amniotic membrane transplantation (AMT), and three had a history of both corneal transplantation and AMT. The interval between the occurrence of the patients’ chemical burns and enrollment in this study varied from 1 to 180 months (mean = 33.27 ± 51.79 months). According to the medical histories and physical examinations, five patients had developed corneal neovascular vessels within a month after the chemical burn (Table 1. Situation, new), and the other six patients had CNV more than one month after the chemical burn (Table 1. Situation, old). The detailed patient characteristics are listed in Table 1.

Treatment with Subconjunctival Injections of Bevacizumab and Follow-up

All eyes were administered a subconjunctival injection of bevacizumab after enrollment. Before injection, the ocular surface was anesthetized using proparacaine hydrochloride (Alcon Inc., Fort Worth, TX, USA). Subsequently, bevacizumab was injected subconjunctivally at a site adjacent to the entrance of the CNV. A dose of 2.5 mg/0.1 mL per involved corneal quadrant was injected to form a conjunctival bulla. After the injection, each patient was treated with 0.3% tobramycin/0.1% dexamethasone eye drops (TobraDex, Alcon Inc.) four times a day for four weeks. All eyes received a second subconjunctival injection of bevacizumab in the fourth week after the first injection. After the second injection, tobramycin/dexamethasone eye drops were administered four times a day for the next four weeks.

Best-corrected visual acuity (BCVA) and intraocular pressure (IOP) were measured regularly, and slit-lamp biomicroscopy and digital photography were performed before and after treatment at different follow-up visits. IOP was measured using Goldman applanation tonometry. Patients visited at weeks 2, 4, 6, 24, and 48 after the first injection of bevacizumab. 

Evaluation of Corneal Neovascularization 

Corneal photographs were used to evaluate morphological changes in the corneal neovascular vessels. At each visit, at least three front-view corneal photographs (with the light source at an angle of 15° on the left side of the patient) were obtained at magnifications of 10× and 16× using a digital camera (Canon DS126291, Canon Inc., Tokyo, Japan) attached to a slit-lamp microscope (Topcon SL-7E, Topcon Corporation, Tokyo, Japan). 

The digital slit-lamp corneal photographs were analyzed by a single masked observer using two graphic software applications; Photoshop® CS2 (Adobe Systems Inc., San Jose, CA, USA) and Image-Pro Plus 4.0 (Media Cybernetics, Silver Spring, MD, USA) for Microsoft Windows XP. As described in a previous report,9 the total corneal area was first delineated using Photoshop, and then the corneal vasculature was manually delineated using the computer’s cursor. The area of CNV and the length and diameter of the neovascular vessels were measured using Image-Pro Plus (Figure 1). The neovascular area (NA) was defined as the area occupied by neovascular vessels and not the involved corneal area. Neovascular length (NL; the approximate cumulative length of all vessels) and neovascular diameter (ND, the mean diameter of all vessels) were obtained by measuring the approximate cumulative length and the mean diameter of all the corneal vessels, respectively.9 When determining the NL and ND, only the longest branch and the widest branch were measured for one vessel tree.

Statistical Analysis

Statistical analysis was performed by Professor Fu-tian Luo from the Department of Biomedical Statistics of the School of Public Health at Sun Yat-sen University. Because of the unknown distribution of the observed data and the small sample size in this study, statistical analysis was performed using the R package nparLD for Microsoft Windows XP, which implements a broad range of rank-based nonparametric methods for analyzing longitudinal data in factorial experiments. The LD-F1 model was applied to compare the whole dataset at different checkpoints. The F1-LD-F1 model was used to compare the changes in new (developed within the first month) and old (noted more than one month ago, and vessels with fibrovascular membranes) vessels.16 A P value of less than 0.05 was considered statistically significant. A modified P value (modified ANOVA-type statistic) was used for comparisons between two groups, defined as 0.05/the comparison times.

Results

The patients’ information before and after treatment, is summarized in Table 1. All patients were followed for one year after bevacizumab injection. At the last visit, the BCVA was found to have improved in five patients (from 1 to 5 lines), remained unchanged in five patients, and decreased in one patient compared to pretreatment.

Changes in Corneal Neovascular Vessels 

The values of the CNV parameters after treatment were converted into reduced percentages of the baseline (before treatment) values. All data in this study were described using three box plots (Figure 2), and these indicated that the measured distances had skewed distributions. An increase in the median gave rise to a time effect. Compared with the baseline values, the three parameters of area (NA, a and d), length (NL, b and e), and diameter (ND, c and f) decreased significantly (< 0.001) at all checkpoints. By observing the plot along with the pointwise 95% confidence intervals, it was found that the percentages of CNV improvement increased over time in both groups (Figure 2a–c). Multiple comparisons were performed to investigate the effect between the two contiguous groups. The reduced percentages of NA increased until six months after the first injection (24–48 weeks, p = 0.27). However, NL and ND reductions peaked at four weeks after the first injection (4-6 weeks, for NL = 0.14, for ND = 0.13).

A similar statistical method was used to analyze the different effects between the new and old neovascular vessel groups; the F1-LD-F1 model in the R package was used in this case. Box plots indicated that the measured distances of the three parameters in both groups had somewhat skewed distributions, and the relative effects increased gradually over time. The regression of the new neovascular vessels in five patients was significantly greater than that of the older neovascular vessels in the other six patients either after the first or second injection, and both groups exhibited a time effect in Figure 2d–f (NA, NL, ND, group and time < 0.05). When NA and NL were compared across two contiguous time groups, it was found that the effect of the two groups increased over time until six months (24–48 weeks, time NA p = 0.225; NL p = 0.056). The ND decreased until six weeks after treatment (6–24 weeks, = 0.134). The regression NA of the new and old neovascular vessels showed a statistically significant difference only after the first injection (0–2 weeks, group p < 0.001, other > 0.05). When the NL was examined, it was found that the effects of the two groups were similar at all time points (> 0.05). The regression ND of the two groups was statistically different until four weeks (0–2, 2–4 weeks, group p < 0.001, other > 0.05).

Among the five patients who had newly formed CNVs, two patients developed neovascular vessels at the interface between the lamellar corneal graft and native corneal stroma within one month after anterior deep lamellar keratoplasty. Four weeks after treatment with subconjunctival bevacizumab, the corneal vessels regressed considerably (Figure 2a–c). Three patients with a history of chemical burns in the previous month exhibited abundant superficial vessels in the peripheral cornea. After treatment, the superficial corneal vessels regressed significantly, and the vessels in the fibrovascular membrane narrowed (Figure 2f). Meanwhile, the corneal edema improved. All patients showed improved BCVA. 

In this study, six patients had corneal neovascular vessels (with fibrovascular membranes) that were more than one month old. The large vessels were slightly shorter and narrower than before treatment in Case 7 (Figure 3 A–C). Meanwhile, the number of quadrants involved in the CNV did not change. Most of the fine neovascular vessels, which could previously be identified with slit-lamp microscopy, became invisible after treatment as in Case 10 (Figure 3 D–F). In Case 2, corneal lipid deposition was slightly lower after bevacizumab treatment (Figure 3G and 3H).

Complications 

In Case 1, the corneal epithelial defect areas did not increase or improve after bevacizumab treatment. Filamentary keratitis was observed after treatment in Cases 2, 3, and 9. After the removal of the small white filaments, there was no relapse of 

filamentary keratitis. In Case 2, an inferior corneal epithelial defect was noted four weeks after injection. Trichiasis of the upper eyelid was noted in this patient. The epithelial defect healed after the trichiasis was cleared, the patient wore a therapeutic corneal contact lens (Purevision®, Bausch & Lomb Incorporated, Rochester, NY, USA), and the tobramycin/dexamethasone eye drops were terminated.

All patients, except Case 4, had normal IOP during follow-up. The IOP of Case 4’s left eye was 35.7 mmHg (measured using non-contact tonometry) two weeks after treatment. After discontinuation of topical tobramycin/dexamethasone eye drops for four weeks, the IOP decreased to 11.7 mmHg. No systemic adverse events were observed.

Discussion

This study has shown that bevacizumab treatment could reduce the length and diameter of neovascular vessels that develop secondary to corneal burns. The peak effect on CNV occurred four weeks after bevacizumab treatment. A probable reason for this is that the bevacizumab concentration in the normal cornea stroma peaked for 14 days and began to decline at 21 days after the subconjunctival injection.17 

The most interesting finding in this study was that subconjunctival administration of bevacizumab successfully reduced newly formed CNV. This is in alignment with the findings of other studies. Partial regression of CNV after subconjunctival injection of bevacizumab was also observed in high-risk transplanted corneas,20 in a rabbit limbal insufficiency model,18 and in a rabbit alkali burn-induced CNV model.19 

Cursiefen et al. found that more than 80% of new vessels were covered by pericytes within two weeks of CNV onset.18 Therefore, the optimal time to administer bevacizumab to induce regression of the immature neovascular vessels is no later than when pericyte coverage occurs. However, the effects of bevacizumab on CNV with fibrovascular tissues, such as primary,21 recurrent11, or impending recurrent pterygium,22 are limited. In these situations, topical use of bevacizumab could only partially and transiently decrease conjunctival vascularization,22 not CNV.11 A recent review, without direct evidence, recommended early treatment with anti-angiogenic agents for preventing CNV and formation of fibrovascular pannus in patients with chemical burns.23 The results presented here and in our previous report24 support this recommendation. In our previous report, early application of subconjunctival bevacizumab significantly prevented CNV after sclerocorneal lamellar keratoplasty for chemical burns.24

In this study, CNV that developed within a month mostly regressed after treatment. However, vessels with fibrovascular membranes (that formed more than one month ago) were observed to be narrower and shorter compared to pretreatment. Although the improvement in these older cases of CNV was statistically significant, it may not be clinically significant because the quadrants with CNV involvement were unchanged. Therefore, the risk of corneal graft rejection remains high. Moreover, it was noted that bevacizumab treatment did not inhibit pseudopterygium growth; however, itb did reduce the severity of CNV with fibrovascular membrane formation. As demonstrated in previous reports, alleviation of lipid deposition may contribute to narrowing of large vessels8 and reduced vascular leakage25-27 after bevacizumab treatment. 

An increase in blood pressure is commonly reported after systemic administration of bevacizumab for rectal cancer.28 However, no systemic events associated with topical eye drops and subconjunctival injection of bevacizumab have been reported,8 and none were observed in this study. Punctuate epithelial erosion,29 new corneal epithelial defects,30 and corneal thinning have been reported in the second month after topical application of bevacizumab.31 In this study, filamentary keratitis was noted in three patients and a new epithelial defect in one patient. Filamentary keratitis may be related to ocular surface instability, and the new epithelial defect was attributed to eyelid trichiasis rather than an adverse effect of bevacizumab, which is consistent with the findings of a previous study.32 It has yet to be determined whether injected bevacizumab is toxic to the corneal epithelium. Various conflicting results have been reported in different in vitro,33,34 and in vivo animal studies,35,36 as well as in human applications.37 

The main drawbacks of this study were the lack of uniformity among the enrolled patients, the small sample size, and the follow-up period being insufficient for long-term evaluation. However, there has been no report to date on the behavioral differences between stromal and superficial corneal vessels. Well-controlled prospective clinical trials are required to determine the optimal dose, route of administration, frequency of use, and possible side effects of this treatment. An innovative method for visualizing and quantifying both superficial and deep  corneal neovascular vessels is needed to capture the effects on all corneal vessels38 .

In conclusion, this study has shown that subconjunctival injection of bevacizumab could decrease the severity of CNV that develops secondary to chemical burns, especially in newly formed vascular vessels. It is recommended that bevacizumab be applied as soon as possible after corneal chemical burns to prevent CNV is recommended as soon as possible after stabilization of ocular surface epithelialization. Corneal epithelial defects should be closely monitored, and drugs for corneal epithelialization are necessary.

Declarations

Acknowledgement: This study was supported by the Fundamental Research Funds of the State Key Laboratory of Ophthalmology (Grant number: 303060202400201119 in China.

Author contribution statement: Conceptualization, Shi-you Zhou; data acquisition and/or research execution methodology, Wen-yan Peng, Xiao-fang Yin, Binbing Zhou, and Tao Zhou; data analysis and/or interpretation, Wen-yan Peng and Xiao-fang Yin; manuscript preparation, Wen-yan Peng and Xiao-fang Yin; and manuscript review, Shi-you Zhou. All authors have read and agreed to the published version of the manuscript.

Additional Information

Declaration of competing interest: The authors have no competing interests to declare that are relevant to the content of this article.

Data availability

The datasets analysed during the current study available from the corresponding author on reasonable request.

References

  1. Bachmann B, Taylor RS, Cursiefen C. Corneal Neovascularization as a Risk Factor for Graft Failure and Rejection after Keratoplasty An Evidence-Based Meta-analysis. Ophthalmology 2010;117:1300-U42.
  2. Nakao S, Hata Y, Miura M, et al. Dexamethasone inhibits interleukin-1beta-induced corneal neovascularization: role of nuclear factor-kappaB-activated stromal cells in inflammatory angiogenesis. Am J Pathol 2007;171:1058–65.
  3. Nirankari VS. Laser photocoagulation for corneal stromal vascularization. Trans Am Ophthalmol Soc 1992;90:595–669.
  4. Pillai CT, Dua HS, Hossain P. Fine needle diathermy occlusion of corneal vessels. Invest Ophthalmol Vis Sci 2000;41:2148–53.
  5. Yoon KC, You IC, Kang IS, et al. Photodynamic therapy with verteporfin for corneal neovascularization. Am J Ophthalmol 2007;144:390–5.
  6. Cursiefen C, Cao J, Chen L, et al. Inhibition of hemangiogenesis and lymphangiogenesis after normal-risk corneal transplantation by neutralizing VEGF promotes graft survival. Invest Ophthalmol Vis Sci 2004;45:2666–73.
  7. Carrasco MA. Subconjunctival bevacizumab for corneal neovascularization in herpetic stromal keratitis. Cornea 2008;27:743–5.
  8. Chu HS, Hu FR, Yang CM, et al. Subconjunctival injection of bevacizumab in the treatment of corneal neovascularization associated with lipid deposition. Cornea 2011;30:60–6.
  9. You IC, Kang IS, Lee SH, et al. Therapeutic effect of subconjunctival injection of bevacizumab in the treatment of corneal neovascularization. Acta Ophthalmol 2009;87:653–8.
  10. Uy HS, Chan PS, Ang RE. Topical bevacizumab and ocular surface neovascularization in patients with stevens-johnson syndrome. Cornea 2008;27:70–3.
  11. Bahar I, Kaiserman I, McAllum P, et al. Subconjunctival bevacizumab injection for corneal neovascularization in recurrent pterygium. Curr Eye Res 2008;33:23–8.
  12. Doganay S, Firat PG, Cankaya C, et al. The subconjunctival administration of bevacizumab inhibits corneal neovascularization effectively in the rabbit corneal alkali burn model. Burns 2013;39:326–30.
  13. Gal-Or O, Livny E, Sella R, et al. Efficacy of Subconjunctival Aflibercept Versus Bevacizumab for Prevention of Corneal Neovascularization in a Rat Model. Cornea 2016;35:991–6.
  14. Eiger-Moscovich M, Livny E, Sella, R, et al. Comparison of Subconjunctival Aflibercept and Betamethasone for the Treatment of Formed Corneal Neovascularization in a Rabbit Model. Ophthalmic Res 2019;62:116–122.
  15. Dekaris I, Gabrić N, Drača N, et al. Three-year corneal graft survival rate in high-risk cases treated with subconjunctival and topical bevacizumab. Graefes Arch Clin Exp Ophthalmol 2015;253: 287–294.
  16. Frank Konietschke, EB, Yulia R. Gel, et al. nparLD: An R Software Package for the Nonparametric Analysis of Longitudinal Data in Factorial Experiments. J Stat Softw 2012;50:1–23.
  17. Dastjerdi MH, Sadrai Z, Saban DR, et al. Corneal penetration of topical and subconjunctival bevacizumab. Invest Ophthalmol Vis Sci 2011;52:8718–23.
  18. Lin CT, Hu FR, Kuo KT, et al. The different effects of early and late bevacizumab (Avastin) injection on inhibiting corneal neovascularization and conjunctivalization in rabbit limbal insufficiency. Invest Ophthalmol Vis Sci 2010;51:6277–85.
  19. Papathanassiou M, Theodossiadis PG, Liarakos VS, et al. Inhibition of corneal neovascularization by subconjunctival bevacizumab in an animal model. Am J Ophthalmol 2008;145:424–31.
  20. Vassileva PI, Hergeldzhieva TG. Avastin use in high risk corneal transplantation. Graefes Arch Clin Exp Ophthalmol 2009;247:1701–6.
  21. Enkvetchakul O, Thanathanee O, Rangsin R, et al. A randomized controlled trial of intralesional bevacizumab injection on primary pterygium: preliminary results. Cornea 2011;30:1213–8.
  22. Lekhanont K, Patarakittam T, Thongphiew P, et al. Randomized controlled trial of subconjunctival bevacizumab injection in impending recurrent pterygium: a pilot study. Cornea 2012;31:155–61.
  23. Cursiefen C, Colin J, Dana R, et al. Consensus statement on indications for anti-angiogenic therapy in the management of corneal diseases associated with neovascularisation: outcome of an expert roundtable. Br J Ophthalmol 2012;96:3–9.
  24. Huang ST, Zhou T, Yang YX, et al. Early Application of Bevacizumab After Sclerocorneal Grafting for Patients With Severe Late-Stage Ocular Chemical Burns. Cornea. 2020;39(6):754–760.
  25. Ferrara N. Role of vascular endothelial growth factor in regulation of physiological angiogenesis. Am J Physiol Cell Physiol 2001;280:C1358-66.
  26. Avery RL, Pieramici DJ, Rabena MD, et al. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006;113:363–72 e5.
  27. Lazic R, Gabric N. Intravitreally administered bevacizumab (Avastin) in minimally classic and occult choroidal neovascularization secondary to age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol 2007;245:68–73.
  28. Schmid-Kubista KE, Krebs I, Gruenberger B, et al. Systemic bevacizumab (Avastin) therapy for exudative neovascular age-related macular degeneration. The BEAT-AMD-Study. Brit J Ophthalmol 2009;93:914–9.
  29. Jeong JH, Chun YS, Kim ES, et al. Compensatory growth factor and cytokine response in tears after subconjunctival bevacizumab injection. Cornea 2011;30:1071–7.
  30. Koenig Y, Bock F, Horn F, et al. Short- and long-term safety profile and efficacy of topical bevacizumab (Avastin) eye drops against corneal neovascularization. Graef Arch Clin Exp 2009;247:1375–82.
  31. Kim SW, Ha BJ, Kim EK, et al. The effect of topical bevacizumab on corneal neovascularization. Ophthalmology 2008;115:e33-8.
  32. Bock F, König Y, Kruse F, et al. Bevacizumab (Avastin) eye drops inhibit corneal neovascularization. Graefes Arch Clin Exp Ophthalmol 2008;246:281–4.
  33. Chalam KV, Agarwal S, Brar VS, et al. Evaluation of cytotoxic effects of bevacizumab on human corneal cells. Cornea 2009;28:328–33.
  34. Yoeruek E, Spitzer MS, Tatar O, et al. Safety profile of bevacizumab on cultured human corneal cells. Cornea 2007;26:977–82.
  35. Mello GR, Pizzolatti ML, Wasilewski D, et al. The effect of subconjunctival bevacizumab on corneal neovascularization, inflammation and re-epithelization in a rabbit model. Clinics (Sao Paulo) 2011;66:1443–50.
  36. Kim TI, Chung JL, Hong JP, et al. Bevacizumab application delays epithelial healing in rabbit cornea. Invest Ophthalmol Vis Sci 2009;50:4653–9.
  37. Koenig Y, Bock F, Horn F, et al. Short- and long-term safety profile and efficacy of topical bevacizumab (Avastin) eye drops against corneal neovascularization. Graefes Arch Clin Exp Ophthalmol 2009;247:1375–82.
  38. Faraj LA, Said DG, Dua HS. Evaluation of corneal neovascularisation. Br J Ophthalmol 2011;95:1343–4.

Table

Table 1 is available in the Supplementary Files section