Nummular infiltrates in Acanthamoeba keratitis.

Abstract

Sir, Radial keratoneuritis (Moore et al. 1986) and ring-shaped infiltrates (Illingworth & Cook 1998) are regarded as typical clinical features in Acanthamoeba keratitis. Epithelial irregularities (Cohen et al. 1987; Bacon et al. 1993) and stromal infiltrates (Illingworth & Cook 1998) are common, but less typical. Dense subepithelial infiltrates were described by Holland et al. (1991). Here, we present a patient with Acanthamoeba keratitis showing subepithelial infiltrates similar to nummular infiltrates in adenoviral keratitis. Our patient was a 22-year-old woman who wore soft contact lenses. She had already suffered from very painful keratitis of the left eye for 2 weeks when she presented at the clinic for the first time. Slit-lamp examination revealed irregularities of the epithelium and keratoneuritis (Fig. 1). The anterior chamber was quiet. The right eye had no history of conjunctivitis or keratitis. From the clinical point of view, Acanthamoeba keratitis seemed to be the most probable diagnosis. Left cornea of the 22-year-old female patient with keratoneuritis (arrows). A 200-µl sample from the patient's contact lens container was collected aseptically. The suspension was spotted on non-nutrient agar (NNA) plates (1.5% plain agar in aqua dest) seeded with live Klebsiella aerogenes and incubated at 20 °C. The NNA plates were examined daily for trophozoites and cysts. Amoebic isolates were identified by morphological criteria and by 18S rRNA in situ hybridization. The in situ hybridization was performed with an Acanthamoeba genus-specific oligonuceleotide probe as described previously (Stothard et al. 1999; Grimm et al. 2001). Furthermore, material harvested from the corneal surface was analysed in the same way. Isolation and microscopic analysis of protozoa revealed that Acanthamoeba trophozoites and cysts were present in the liquid in the contact lens container. The positive in situ hybridization signals confirmed the prior identification by morphological criteria. The optimal hybridization signals were obtained with 25% formamid. Using these stringent conditions, the probe specifically hybridized the Acanthamoeba isolates as well as Acanthamoeba reference strains (Grimm et al. 2001). We were unable, however, to detect Acanthamoeba species in the material harvested from the cornea. Kanamycin sulfate (0.65%; Alcon Pharma, Freiburg, Germany), polyhexamethylene biguanide (0.02%; own pharmacy) and propamidine isethionate (0.1%; May & Baker Ltd, Dagenham, Essex, UK) eyedrops were administered every 15 min for the next 10 days. Thereafter, these eyedrops were tapered slowly. Topical prednisolone 1% acetate application (Pharm-Allergan GmbH, Ettlingen, Germany) was started 6 days after referral to the clinic, with a daily dose of two drops. After 1 month, all drugs (kanamycin sulfate, polyhexamethylene biguanide, propamidine isethionate and prednisolone acetate) were administered three times daily. At this time subepithelial infiltrates similar to nummular infiltrates in adenoviral keratitis were observed for the first time (Fig. 2). Because nummular infiltrates in adenoviral keratitis are known to respond to topical cyclosporin A (Reinhard et al. 2000), topical cyclosporin A 2% (own pharmacy) was administered twice daily and therapy with prednisolone 1% acetate was stopped. With this regimen the infiltrates disappeared slowly over the following 4 months (Fig. 3). Topical therapy with kanamycin sulfate, polyhexamethylene biguanide and propamidine isethionate was stopped after 12 months. The patient is still receiving cyclosporin A eyedrops once daily 12 months after onset of the keratitis without any sign of recurrence. Subepithelial infiltrates similar to nummular infiltrates in adenoviral keratitis after 1 month. Resolution of all inflammatory signs 4 months after onset of the keratitis. In this case, the diagnosis of Acanthamoeba keratitis was made via a typical history (soft contact lens wearer, very painful keratitis), detection of keratoneuritis at the slit-lamp and detection of Acanthamoeba species via immuno-genetic examination of the soft contact lens container. Subepithelial infiltrates were observed for the first time after 4 weeks. They were less dense than those described by Holland et al. (1991) and resembled nummular infiltrates in adenoviral keratitis. The subepithelial infiltrates described by Holland et al. (1991) in six patients occurred up to 18 months after onset of Acanthamoeba keratitis, mostly when the patients were still receiving topical corticosteroids. In our case, the infiltrates disappeared when topical cyclosporin A was administered. Most probably, the infiltrates were caused by an immunological reaction against Acanthamoeba antigen in the anterior corneal stroma. This is the first report of subepithelial infiltrates similar to nummular infiltrates in adenoviral keratitis in a patient with Acanthamoeba keratitis. The infiltrates disappeared with combined therapy of topical anti-amoebic agents and cyclosporin A.