Dendritic keratitis is a rare complication following varicella infection in children.1 5 Typically, the patient has a stromal disciform keratitis which occurs following the cutaneous eruption. This may occur anywhere from 1 to 13 weeks after the onset of the rash.4·5 The development of a dendrite may then occur during the course of the stromal keratitis. The clinical course is often self-limited6,7; however, patients may require cycloplegics and topical steroids, particularly if there is an associated iridocyclitis. We encountered a case of varicella stromal disciform keratitis complicated by dendrite formation, which progressed despite therapy and resulted in stromal scarring, irregular astigmatism, and amblyopia.
A 5-year-old boy developed a typical cutaneous varicella eruption which spontaneously resolved. One month later, he developed a mild irritation of his left eye but was otherwise asymptomatic. This worsened over a period of 3 months and he was ultimately seen by an ophthalmologist with chief complaints of pain, redness, and photophobia. On examination, he was noted in the left eye to have decreased vision to light perception, corneal stromal edema, a 4 + cellular anterior chamber reaction (greater than 50 cells per high power field), and an elevated intraocular pressure (IOP) of 35 mm Hg. The right eye was normal.
The referring ophthalmologist diagnosed herpes simplex virus (HSV) keratitis and iridocyclitis. The patient was placed on a regimen of trifluridine four times a day (qid), betaxolol 0.50% twice a day, prednisolone acetate 1% qid, and scopolamine 0.25% three times a day to the affected left eye. The visual acuity improved to 20/60 with a decreased anterior chamber cellular reaction (20 to 50 cells per high power field); however, the patient developed a new epithelial keratitis and endothelial keratic precipitates over the following 3 weeks. The IOP returned to normal and the boy was referred to us for evaluation of a persistent keratouveitis.
On examination, the visual acuity was OD: 20/20, OS: 20/100, pinhole 20/80. There was decreased corneal sensation of the left eye with a disciform keratitis centrally. Central corneal thickness was measured with an ultrasonic pachymeter and was OD: 550 pm; OS: 580 µp?. There was a dendriform lesion centrally which was elevated and stained moderately with fluorescein (Fig 1). Corneal culture was negative for HSV and serum antibody was negative for HSV. The patient was begun on treatment with acyclovir ointment (3%) four times daily and topical prednisolone acetate four times daily. One week later, there was no improvement in the patient's condition and the dendrite persisted (Fig 2). The dosage of prednisolone acetate 1% was increased to six times daily and acyclovir therapy was unchanged. Three weeks later, the dendrite had resolved and the corneal edema was greatly improved, with central corneal thickness decreased to 550 µp?. Although the central cornea had significantly less opacity, the best-corrected visual acuity remained at 20/100. This confirmed our suspicion that the patient had developed amblyopia. Occlusion patching therapy was begun in the right eye part time (4 to 6 hours daily). The steroid dosage was tapered over a period of 6 weeks to one drop a day and the acyclovir therapy was discontinued.
Four-and-a-half months later, the patient showed total resolution of corneal edema with a central corneal thickness of 510 µp?. However, there was anterior stromal scarring which involved the visual axis (Fig 3A). The patient's visual acuity had improved to 20/50, although he had developed some irregular astigmatism as seen on photokeratoscopy (Fig 3B). Retinoscopy was piano ? ( + ) 1.50 cylinder axis 170°. Eye glasses were prescribed, and the patient was continued on prednisolone acetate 1% one to two times daily, and given instructions for full-time occlusion therapy of the right eye. Follow up 6 weeks later showed an improved visual acuity in the left eye to 20/40 with the right eye remaining at 20/20. The patient was instructed to continue part-time patching of the right eye. Three weeks later, he returned with a recurrence of the dendrites (Fig 4). Corneal scraping of the dendrites failed to reveal the presence of giant cells. Herpes zoster culture and herpes zoster fluorescent antibody were negative. The patient was placed on 10% acetylcysteine four times daily, which led to resolution of the dendrites. His treatment consists of prednisolone acetate Ve% once a day, and he maintains 20/40 visual acuity with part-time occlusion therapy of the contralateral eye.
FIGURE 1: Central disciform keratitis of the left eye with a dendrite. The central pachymetry was 580 am.
FIGURE 2: One week later, the dendrite is slightly changed in appearance but still stains with fluorescein.
FIGURE 3: (A) Five months later, there is central corneal stromal scarring. The edema has resolved and the pachymetry is now 510 µ/?. The epithelium does not stain. (B) Photokeratoscopy reveals irregular mires of the left cornea demonstrating irregular astigmatism.
Dendritic keratitis is an extremely uncommon ocular complication which follows a chickenpox infection. A recent review of the literature found a total of 32 reported cases of varicella disciform stromal keratitis with only nine complicated by dendritic keratitis.5 Few studies have been performed to evaluate the incidence of ocular complications following chickenpox infection and those that have been performed have not shown a high incidence of disciform or dendritic keratitis. In a prospective survey of 82 patients with an acute chickenpox infection, the incidence of a low-grade iritis with or without a punctate keratitis was fairly common, occurring in 25% of patients.8 However, in all of these patients, the ocular inflammation resolved rapidly without therapy. In a series of 24 patients specifically referred with ocular involvement, the most common findings were lid lesions followed by limbal conjunctival pocks.9 In addition, 6 (25%) of 24 patients had a mild or moderate uveitis of which 5 were treated with topical steroids. No patient required therapy for longer than 2 weeks, and all episodes resolved. Two patients had corneal findings, one with punctate keratitis and another with a limbal anterior infiltrate. Both cases resolved without therapy.
Varicella dendritic keratitis is primarily a clinical diagnosis based on the findings of a disciform keratitis followed by dendrite formation in a patient with a recent chickenpox infection. Although microdendrites can appear during either an acute varicella or zoster infection, this appears to be an extremely uncommon finding.6,10 More commonly, dendrites occur later in the course of the infection in both primary varicella and herpes zoster ophthalmicus infections.4,5,11 These dendrites often are called pseudodendrites because their appearance is different from the true dendrites seen in herpes simplex infections. Pseudodendrites typically are whitish-gray, elevated plaques which contain abnormal heaped-up epithelium. They stain moderately well with fluorescein and very brightly with rose bengal. On the other hand, true dendrites are excavated epithelial defects with discrete terminal bulbs. The base of the dendrite stains very brightly with fluorescein and the edges of the dendrite stain only moderately well with rose bengal.
In contrast to the early microdendrites in zoster from which viable virus has been cultured,10 the pseudodendrites which develop later (also called mucous plaque keratitis) do not appear to contain viable virus.11 However, in varicella dendritic keratitis, both viral particles1 and viral antigen3 have been demonstrated in the pseudodendrites. In our patient, viral culture and fluorescent antibody stains were performed on scrapings of the dendrite that appeared at 14 months. Both tests failed to demonstrate the presence of varicella virus.
FIGURE 4: Fourteen months following the initial episode of chickenpox, there is a recurrence of the dendrite superonasally which stains with rose bengal. Visual acuity remains 20/40.
In addition to performing tests for the presence of varicella virus, these patients also need to have an evaluation for the presence of HSV since this is another common cause of disciform keratitis.4,5 Serum antibody titers if negative are very helpful in excluding herpes simplex as a cause of keratitis. Corneal scrapings of the dendrite looking for giant cells and herpes simplex viral culture also are useful. All of these were negative in our patient and helped establish that varicella was the cause of the keratitis.
Topical corticosteroid therapy in patients with varicella disciform keratitis is controversial, with some authors recommending its use while others avoid it.4"6 Some contend that since the disease is self-limited, steroids are not only unnecessary but may encourage relapses and recurrences.7 Although there is no strong evidence suggesting that topical steroids either prolong or alter the course of inflammation, it is interesting to note that the vast majority of reported cases of dendritic formation have been in patients who received steroids.5 Also, once a patient is receiving steroids, it is often difficult to taper the dosage without recurrence of the inflammation.
Our patient required topical steroids early in the course of his illness to control uveitic glaucoma. Subsequently, we continued steroids because of his extremely poor vision due to central corneal edema and our concern that he was at risk for developing irreversible amblyopia. Once the edema had cleared significantly, we began occlusion therapy of the contralateral eye, which improved his vision from 20/100 to 20/40.
There is evidence that systemic acyclovir is beneficial in acute varicella infections in children12; however, ocular disease has not been assessed with either topical or systemic antivirale. In ophthalmic zoster, topical antivirale have not been shown to be of benefit in treating dendrites (mucous plaque keratitis), although 10% acetylcysteine may be effective.11 In our patient, topical acyclovir appeared to have little effect, although the acetylcysteine appeared helpful.
In conclusion, we point out that not all patients who develop disciform keratitis (with or without dendrite formation) have a self-limited clinical course. Uveitic glaucoma may mandate the use of medications to control inflammation and IOR In addition, patients should be monitored closely for the development of form deprivation amblyopia which may require occlusion patching of the contralateral eye. With careful attention, complications that could damage vision may be avoided.
1. Nesburn AB, Borit A, Pentelei-Molnar J, Lázaro R. Varicella dendritic keratitis. Invest Ophthalmol. 1974;13:764-770.
2. Tessler HH, Krimmer BM. Disciform keratitis induced by varicella virus. Eye, Ear, Nose, and Throat Monthly. 1975;54:311.
3. Uchida Y, Kaneko M, Hayashi K. Varicella dendritic keratitis. Am J Ophthalmol. 1980;89:259-262.
4. Uchida Y. Varicella dendritic keratitis. In: Darrell RW, ed. Viral Diseases of the Eye. Philadelphia, Pa: Lea & Febiger; 1985:73-77.
5. Wilmhelmus KR, Hamill MB, Jones DB. Varicella disciform stromal keratitis. Am J Ophthalmol. 1991;111:575-580.
6. DeLuise VP, Wilson FM II. Varicella and herpes zoster ophthalmicus. In: Duane TD, Jaeger EA, eds. Clinical Ophthalmology. Philadelphia, Pa: JB Lippincott Company; 1988;(vol 4) chapter 20:1-21.
7. Tessler HH, Krimmer BM. Corticosteroids and varicella disciform keratitis. Am J Ophthalmol. 1980;90:115-116.
8. Kachmer ML, Annable WL, Dimarco M. Iritis in children with varicella. J Pediatr Ophthalmol Strabismus. 1990;27:221-222.
9. Jordan DR, Noel LP, Clarke WN. Ocular involvement in varicella. Clin Pediatr. 1984;23:434-436.
10. Pavan-Langston D, McCulley J. Herpes zoster dendritic keratitis. Arch Ophthalmol. 1973;89:25-29.
11. Marsh RJ, Cooper M. Ophthalmic zoster: mucous plaque keratitis. Br J Ophthalmol. 1987;71:725-728.
12. Dunkle LM, Arvin AM, Whitley RJ, et al. A controlled trial of acyclovir for chickenpox in normal children. N Engl J Med. 1991;325:1539-1544.