Treatment Pulse

Focus on Herpetic Keratitis

Insights to help you diagnose and treat herpes simplex virus

Special section sponsored by Bausch + Lomb

Eric Donnenfeld, MD and Edward Holland, MD

Dr. Holland: Herpes simplex virus (HSV-type 1 and HSV-type 2) is a common cause of infection. Approximately 25% of the population is seropositive for HSV at age 4, and 100% of people are seropositive at age 60.¹ First exposure to the virus often occurs in childhood and causes a mild or subclinical reaction, after which the virus becomes latent in the trigeminal ganglion or cornea. The virus is reactivated years later by stress, UV radiation, hormonal changes or a compromised immune system.

About 1% of individuals with HSV develop ocular manifestations,¹ which means that about 400,000 Americans have had some form of ocular HSV with 20,000 new cases and 28,000 relapses of keratitis each year.¹ More importantly, HSV is the leading cause of infectious blindness in the United States.¹

Diagnosing HSV Keratitis

Dr. Holland: In HSV keratitis, there are four main clinical presentations: infectious epithelial keratitis, neurotrophic keratopathy, stromal keratitis and endotheliitis. Although we see these conditions often, they can be challenging to differentiate from other causes. Mistaken diagnoses and treatments are common due to the variety of presentations.

First, infectious epithelial keratitis can present with four types of lesions: corneal vesicles, dendritic ulcers, geographic ulcers and marginal ulcers. The prodrome to the dendritic ulcer is the eruption of the virus on the corneal surface in vesicles. The vesicles, which contain active virus, coalesce to form a dendritic ulcer. Fluorescein reveals the absence of stain because the epithelium is still intact, but you will see late staining through the abnormal epithelium. The vesicles coalesce to the dendritic ulcer lesion classically with swollen borders, and then the center part of the lesion ulcerates into Bowman's layer. This dendritic ulcer is what most clinicians recognize as the clinical presentation of HSV keratitis.

Infectious Epithelial Keratitis: Dendritic Ulcer

This most common presentation for HSV shows branching ulceration and swollen epithelial borders that contain active virus.

When we see a branching lesion in the epithelium, our knee-jerk reaction can be HSV. However, not every dendrite is an HSV infectious lesion. What other conditions or situations could cause a branching lesion in the epithelium?

Dr. Donnenfeld: Healing after photorefractive keratec-tomy (PRK) or any abrasion can create a pseudo-dendritic suture coming together, and the last thing you want to do is give an antiviral after PRK or an abrasion.

But the most common misdiagnosis I see is a patient who has had herpetic disease and has a ghost dendritic ulcer. Epithelial remodeling is taking place. While the active virus has been killed, clinicians continue to use an antiviral, causing toxicity that feeds upon itself, augmenting that pseudo-dendrite. In managing HSV keratitis, it's important to remember that antiviral therapy should be efficacious but managing toxicity and inflammation play a role in healing, as well.

Dr. Holland: A dendritic lesion may enlarge, becoming a geographic ulcer, also with active virus in the swollen borders. We see dendritic and geographic lesions in both the conjunctiva and the cornea, so it's something to check for in the conjunctiva when patients have recurrent conjunctivitis.

With a paracentral or central dendritic lesion, we don't see many inflammatory cells early on because there are no vessels. But a limbal lesion has an intense uptake of white cells that causes the lesion to look somewhat like a staph marginal ulcer. If topical steroids are erroneously prescribed in the early replicating phase, the virus will progress.

Every dendritic ulcer leads to some form of persistent dendritic epitheliopathy. As the virus is eliminated and the lesion heals, the dendritic shape persists in the epithelium and may take several weeks to resolve. Clinicians often prolong topical antiviral therapy even though the dendritic ulcer is healed. Neurotrophic epitheliopathy has the same appearance and branching pattern as an infectious lesion (a dendritic ulcer).

The third type of lesion and biggest chronic clinical challenge is stromal keratitis. Epidemiological evidence shows that in about 20% of patients, the infectious epithelial lesion progresses into stromal keratitis.² We typically see an infiltrate, secondary neo-vascularization and sometimes an immune ring. The cause of stromal keratitis is thought to be retained viral antigen, which stimulates an immune reaction, rather than an infectious reaction. However, necrotizing stromal keratitis is a rapid necrosis of the cornea attributed to an intense immune reaction to a replicating virus in the stroma. My experience is that these patients need therapeutic keratoplasty quite quickly. Fortunately, this is rare.

HSV Neurotrophic Keratopathy

This patient was 4 weeks into treatment of HSV keratitis with trifluridine (Viroptic). When the branching lesion persisted, the clinician kept treating with trifluridine. However, infectious lesions only last 4 weeks in immuno-compromised patients. The central lesion isn't a geographic lesion. Its edges are smooth, not scalloped. We see punctate epithelial erosions, ulcer and dendritic epitheliopathy. This is a neurotrophic ulcer from the virus' effects on the surface, including toxicity, inflammation and dry eye.

Finally, the last corneal presentation is endotheliitis. The virus is present at the endothelial level, and an immune reaction actually targets the endothelium. We see an associated anterior chamber reaction. These patients actually can have an elevated intraocular pressure from a trabeculitis–sometimes quite high. Endotheliitis can have a round or disciform shape, a linear appearance or a diffuse involvement.

An Antiviral Time Leap

Dr. Donnenfeld: As always, treatment rests upon a few very simple principles: eradicate the active virus, and once it's dead, try to allow normal epithelial wound healing. The toxicity of herpetic disease and the resulting neurotrophic keratitis feed very well into the intrinsic damage caused by the antiviral itself, as well as its preservative. Thus, epithelial toxicity is a rate-limiting factor in the use of many antivirals.

I never prescribe a topical antiviral for more than 3 weeks, and I very rarely prescribe it for more than 2 weeks because the risk of toxicity outweighs the benefits of the drug. In many cases, I have used oral antivirals to avoid surface toxicity when treating ocular infections.

Topical options have been limited over the years. Until 2010, trifluridine (Viroptic), approved more than 30 years ago, was the only option. To put this into perspective, 30 years ago, aminoglycosides were new and we performed intracapsular cataract surgery with 7.0 silk sutures.

Trifluridine is activated by both cell and viral thymidine kinase, which means that it can be toxic to human cells with prolonged use. Combine this nonselective nature with the fact that trifluridine has to be used 9 times a day with a thimerosal preservative and requires refrigeration, and you realize it has very limited use.

Today, we're excited about the advent of an antiviral that is changing the way we address this disease. Ganciclovir 0.15% ophthalmic gel (Zirgan) is indicated for treating dendritic herpetic keratitis. Gancilovir is a targeted purine nucleoside that is selectively activated by the virus DNA. When tri-phosphorylated, it inhibits viral replication by two competitive mechanisms: inhibition of viral DNA polymerase and direct incorporation into the virus, which creates a faulty DNA. This selectivity allows it to preferentially target HSV-infected cells over healthy corneal cells.

Important Risk Information for ZIRGAN

ZIRGAN (ganciclovir ophthalmic gel) 0.15% is indicated for topical ophthalmic use only. Patients should not wear contact lenses if they have signs or symptoms of herpetic keratitis or during the course of therapy with ZIRGAN. Most common adverse events reported in patients were blurred vision (60%), eye irritation (20%), punctate keratitis (5%) and conjunctival hyperemia (5%).

Clinical Studies of Efficacy

Dr. Donnenfeld: Four studies led to FDA approval of ganciclovir. Three were comparative, randomized clinical trials, and one was an open-label trial. All studies evaluated ganciclovir vs. acyclovir ophthalmic ointment 3%, a third-generation antiviral very similar to ganci-clovir, that has gained favor around the world but is not available in the United States.

Trifluridine wasn't evaluated in these trials, since it's not considered the standard of care in Europe. No placebo was used in the studies because of the severity of disease.

When Phase 2 results³ from three of the studies were pooled together, 72% of the ulcers treated with ganciclovir were healed at day 7 versus 69% for acyclovir.

Results of the Phase 3 clinical trial³ that evaluated ganciclovir versus acyclovir in patients with dendritic ulcers showed 77% healed ulcers at 7 days for ganciclovir and 72% for acyclovir.

Epithelial Healing

Dr. Holland: ciclovir and other antivirals is its selectivity to viral DNA. This specificity means ganciclovir preferentially targets viral DNA over host cellular DNA. Therefore, I am able to provide a topical ophthalmic antiviral that is not only effective but safe.

Dr. Donnenfeld: And because it targets viral DNA over host cellular DNA, epithelial healing can occur. Preventing scarring and promoting re-epithelialization is very important in managing herpetic keratitis. To me, nonhealing epithelial defects can be a refractive emergency because they can go on to cause scarring and loss of vision. Just look at PRK–the longer the epithelial defect is present, the greater the resulting scar. The same thing applies to herpetic disease as well. We want to promote healing.

Dr. Holland: If you have long-term, abnormal epitheliopathy, you'll get anterior stromal scarring even with an intact epithelium.

Dr. Donnenfeld: To me, toxicity is the most important factor. I think punctate keratitis is a primary marker for antiviral toxicity. For all four studies on ganciclovir versus acyclovir, the most common adverse events were blurred vision (57.8% vs. 71.3%), eye irritation (20.5% vs. 36.3%), punctate keratitis (6.8% vs. 12.1%) and conjunctival hyperemia (4.3% vs. 3.8%).

I think the 6.8% incidence of punctate keratitis with ganciclovir is a very nice marker for toxicity.

Visual Outcomes

Dr. Donnenfeld: By addressing the dendritic lesions effectively, we're achieving better visual outcomes. HSV keratitis results in stromal reactions in two ways. There's anterior stromal haze, the immediate stromal reaction to a dendritic ulcer, and then there's recurrent immune stromal keratitis. It is important to address the surface infection as soon as it is identified to minimize the risk of active virus progressing to deeper layers of the stroma.

Dr. Holland: I've heard colleagues say, “We have generic products. Why don't we think about saving costs?”

Since the average herpetic patient has multiple episodes, we'd be adding new rounds of antiviral therapy over and over again and that is why we may see conjunctival scarring, punctal stenosis and in some patients, a need for a nasal lacrimal duct surgery.

Dr. Donnenfeld: And ganciclovir versus trifluridine means 5 times a day versus 9 times in the first week, 3 times a day versus 5 times until healed, unrefrigerated versus refrigerated, and BAK versus thimerosal. These are big changes, and that's why ganciclovir is such a revolutionary step forward for us. We can clear the virus with minimal toxicity.

References

1. Holland EJ, Brilakis HS, Schwartz GS. Herpes Simplex Keratitis, in Cornea, J.H. Krachmer, M.J. Mannis, and E.J. Holland (eds.), Mosby Year Book Publishers, St. Louis, 2004.
2. Wilhelmus KR, Falcon MG, Jones BR. Bilateral herpetic keratitis. Br J Ophthalmol. 1981;65(6):385-387.
3. Wilhelmus KR. The treatment of herpes simplex virus epithelial keratitis. Trans Am Ophthalmol Soc. 2000;98:505-532.

Dr. Donnenfeld is a clinical professor of ophthalmology at New York University Medical Center and a partner in Ophthalmic Consultants of Long Island in Rockville Centre, N.Y.
Dr. Holland is a professor of ophthalmology at University of Cincinnati and director of the cornea service at Cincinnati Eye Institute.