Because corneal neovascularization (KNV) is a significant contributor to corneal blindness, and it can be both challenging and frustrating to treat KNV-related lipid keratopathy, we invite corneal specialists to try mitomycin intravascular chemoembolization (MICE) for this condition. We were captivated by the first reports of MICE, suggested by Dean Ouano, MD, so we adopted it early in our practice (May 2021). Since then, we have treated approximately 14 patients with success.

Here, we discuss the rationale for MICE, ideal candidates, the procedure itself, and the postoperative management of patients who undergo this procedure for
KNV-related lipid keratopathy.

Rationale
MICE can serve as a primary or precursor treatment for KNV-related lipid keratopathy.¹ Mitomycin C (MMC) is an alkylating agent that has cytotoxic effects on blood vessels and has been used outside ophthalmology for years.^2,3 Ophthalmologists are likely familiar with using MMC for glaucoma, pterygium, and laser refractive surgeries. MICE is built on the concept of transarterial chemoembolization (TACE) with MMC. TACE has been used outside of ophthalmology, such as in treating hepatocellular carcinoma with targeted destruction of vascular endothelial cells.⁴

Normally, the cornea is immunologically privileged and equally balanced by pro-angiogenic and anti-angiogenic states.² However, disruption to this balance may promote vascular endothelial cell proliferation and KNV. Thus, targeted MMC therapy can be an effective treatment option for KNV.

In fact, in the first KNV-related lipid keratopathy cases by Ouano et al, all 3 had resolution of KNV and lipid keratopathy, and achieved best-corrected visual acuity ≥20/30 up to 1 year after MICE¹ (Figure 1).

In some cases, MICE may not be sufficient as a standalone procedure for visual rehabilitation. In these cases, clinicians may use MICE to induce KNV regression or debulk the stromal opacity (Figure 2). Clinicians should counsel their patients that additional intraocular surgery (e.g., cataract surgery) may be needed after MICE. Though MICE may help improve the corneal opacity, additional surgery or visual rehabilitation strategies (glasses, contact lenses, etc.) may be needed to optimize patients’ visual potential.

As a precursor treatment, MICE may be utilized to treat KNV and “quiet” the cornea in anticipation of future corneal transplant surgery in KNV eyes that have significant stromal scarring that blocks the visual axis.⁵ We recently published our experiences with such a patient.⁶ In these eyes, corneal transplantation (e.g., deep anterior lamellar keratoplasty or penetrating keratoplasty) poses a substantial risk of graft failure and rejection due to persistent KNV.⁵

Ideal Candidates
The ideal candidate for MICE is as follows:

A.  A patient who has lipid keratopathy secondary to KNV encroaching on the visual axis. (Once the visual axis is significantly involved, the prognosis is less favorable.)

B.  A patient who has an underlying etiology controlled (herpes simplex virus/lid margin disease, etc.)

C.  A patient who has a quiet and uninflamed eye.

D.  A patient who has failed a trial of topical steroids (and antivirals if herpes related)

A comprehensive ophthalmologic assessment should be conducted upon appropriately identifying a candidate for MICE. This should include slit lamp photos, corneal topography, corneal tomography, anterior-segment (AS)-OCT, evaluation of corneal sensation, and corneal pachymetry (ultrasonic).

The Procedure
MICE is performed using an ophthalmic operating room microscope, though some surgeons have anecdotally reported favorable outcomes performing MICE at the slit lamp.

A TSK SteriJect thin-walled 33-G needle or JPB 34-G Nanoneedle (JPB-USA) with a partially filled 1.0 cc “Luer slip” (not Luer lock) syringe is filled with MMC at a 0.3 mg to 0.4 mg concentration.

Of note, surgeons should avoid using a 30 g or 32 g needle, as these needles may be too big to cannulate small corneal vessels. Bending the needle at the hub may allow the surgeon to access the vessel(s) of interest more favorably. The surgeon’s non-dominant hand can use toothed forceps to fixate the eye, while creating favorable ergonomics.

The largest bore corneal vessel(s) inside the limbus is identified, and the needle is angled at approximately 15° to cannulate the vessel. An amount of 0.01 mL to 0.05 mL is injected to ensure that a maximum of 0.02 mg is injected. Surgeons should be mindful that enough force is required to ensure MMC reaches both efferent and afferent vessels. Multiple treatments can be given to vessels if needed, resulting in blanching. Following completion, topical antibiotics and steroids are applied.

Subconjunctival or intracameral antibiotics can be given at the surgeon’s discretion. Similarly, some surgeons may choose to administer subconjunctival bevacizumab and/or triamcinolone to augment the MICE procedure. Copious irrigation of the ocular surface with balanced salt solution to ensure rinsing of MMC from the ocular surface at the end of the case is highly recommended to reduce the potential corneal toxicity induced by MMC. A bandage contact lens can be applied if there are concerns of epithelial abrasion.

Due to the recent development of MICE and limited associated literature, there are no reported complications from the procedure.¹ Nevertheless, it’s crucial to highlight that MICE demands technical precision, necessitating a careful focus on the injection site and proper needle positioning (Ouano D, written communication, Feb. 4, 2022). 

Postoperative Management
Evaluation should be conducted 1 day, 1 month, 3 months, and 6 months after the procedure.

Postoperative management should include topical antibiotics and steroids four times daily. Antibiotics can be stopped 7 days to 10 days after the procedure. Steroids can be tapered over one month or at the surgeon’s discretion. Unless contraindicated, oral antivirals should be maintained pre- and postoperatively per the surgeon’s discretion. Also, patients should be advised to maintain lubrication of the ocular surface with non-preserved tears, per the surgeon’s discretion.

Expected findings postoperative days 1 to 7 are blanched/ablated vessels. At postoperative weeks 1 and 2, a “pizza pie sign” representing blood and lipids trapped in corneal stroma is seen (Figure 3). While this may be visually striking, surgeons should be reassured that this is a “normal” postoperative finding after MICE. The intrastromal blood and lipids may take several weeks or months to regress. 

Examinations at postoperative months 1 to 3 may reveal clinical findings consistent with lipid absorption, stromal compaction, and epithelial remodeling.

AS-OCT may be helpful in visualizing stromal compaction, as well as comparison with preoperative tomography. Worsening regular or irregular astigmatism may also be a finding at this time. Visual rehabilitation often requires hard or scleral contact lens usage.

An Intriguing Intervention
In our experience, MICE has been an effective option for the treatment of KNV-related lipid keratopathy. That said, we emphasize that while the results of MICE are full of promise, these are short-term results, so additional, long-term follow-up is needed to determine the role of MICE in the future. Additionally, further research is required to refine the MICE technique, appropriately determine candidate patients, and identify potential long-term outcomes and complications. CP

Neal Rangu, BA is a 3rd year medical student at the University of Oklahoma College of Medicine, Oklahoma City, Oklahoma. He hopes to pursue a career in ophthalmology.

Kamran M. Riaz, MD is an associate professor at the Dean McGee Eye Institute (University of Oklahoma) where he serves as the director of Medical Student Research and director of the Stephenson Refractive Surgery Center. His clinical, surgical, and research practices focus on cornea, external disease, anterior segment, cataract, and refractive surgery.

References

1. Mimouni M, Ouano D. Initial outcomes of mitomycin intravascular chemoembolization (MICE) for corneal neovascularization. Int Ophthalmol. 2022;42(8):2407-2416. doi: 10.1007/s10792-022-02240-6.

2. Seki Y, Toba K, Fuse I, Sato N, Niwano H, Takahashi H, Tanabe N, Aizawa Y. In vitro effect of cyclosporin A, mitomycin C and prednisolone on cell kinetics in cultured human umbilical vein endothelial cells. Thromb Res. 2005;115(3):219-28. doi: 10.1016/j.thromres.2004.09.001.

3. Hoorn CM, Wagner JG, Petry TW, Roth RA. Toxicity of mitomycin C toward cultured pulmonary artery endothelium. Toxicol Appl Pharmacol. 1995 Jan;130(1):87-94. doi: 10.1006/taap.1995.1012.

4. Yamada R, Bassaco B, Bracewell S, Volin S, Collins H, Hannegan C, Guimarares M. Combined conventional transarterial chemoembolization with Mitomycin and percutaneous ablation for unresectable hepatocellular carcinoma. J Gastrointest Oncol. 2020;11(2):298-303. doi: 10.21037/jgo.2019.01.07.

5. 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(7):1300-5.e7. doi: 10.1016/j.ophtha.2010.01.039.

6. Rangu N, Riaz KM. Mitomycin intravascular chemoembolization (MICE) to treat corneal vascularization prior to penetrating keratoplasty. Am J Ophthalmol Case Rep. 2024 Jan 14;33:101993. doi: 10.1016/j.ajoc.2024.101993.