Presumed late recurrence of Acanthamoeba keratitis exacerbated by exposure to topical corticosteroids.

A 28-year-old female with a history of contact lens wear presented with a 1 week history of pain and photophobia in her left eye. In vivo confocal microscopy (IVCM) and corneal scrape confirmed the diagnosis of Acanthamoeba keratitis (AK) which was treated with intensive topical propamidine isethionate (0.1%) and chlorhexidine (0.02%) with tapering dosage over 11 months. Five years after complete resolution of AK and cessation of all contact lens wear, the subject presented to her optometrist with a history of ocular discomfort and mild photophobia. Without further investigation she was prescribed topical corticosteroids. Three weeks later she presented with pain and reduced vision in the left eye. Slit-lamp examination revealed focal, inferior corneal stromal edema. IVCM confirmed widespread Acanthamoeba cysts. Treatment with topical polyhexamethylene biguanide (PHMB) 0.02% and propamidine isethionate 0.1% resulted in resolution of the AK. Despite an initially mild AK, this subject presumably retained viable Acanthamoeba cysts in her cornea 5 years after the initial episode. This report highlights the importance of caution when using corticosteroids in patients with a previous history of AK, even in the relatively distant past. Patients with AK should be warned regarding the risks of recurrence following presumed resolution.

Introduction

Acanthamoeba keratitis (AK) is typically associated with contact lens wear, especially when there is associated poor contact lens hygiene.[1] Numerous studies have highlighted the importance of early diagnosis in ensuring favorable outcomes in AK, however, there are little published data regarding the recurrence rate, the long-term excysting effect of corticosteroid in the acute phase, or the effects of reintroduction of topical corticosteroids years after the acute event.[234]

We report a unique case of markedly delayed recurrence of AK exacerbated by topical corticosteroid use.

Case Report

A 28-year-old female with keratoconus initially presented with a 1 week history of pain and photophobia in her left eye (OS). She was a rigid gas permeable contact lens wearer who admitted to regularly “cleaning” contact lenses in tap water supplied from a water tank. On examination, her best spectacle corrected vision (BSCVA) was 20/20 OD and 20/32 OS. A dendritic-like epithelial lesion was observed on the left cornea and she was treated with oc. acyclovir 5 times daily for presumed herpes simplex keratitis. However, 10 days later she remained symptomatic and a persistent dendritic epithelial lesion with granular subepithelial haze was noted. A diagnosis of AK was suspected and in vivo confocal microscopy (IVCM) confirmed the presence of Acanthamoeba cysts in the basal corneal epithelium OS [Figure 1]. A corneal scrape confirmed the diagnosis of AK.

Figure 1

In vivo confocal microscopy image taken at initial presentation using Confoscan 2. Multiple acanthamoeba cysts are present at the level of the basal corneal epithelium

She was treated with topical propamidine isethionate (Brolene 0.1%) and chlorhexidine (0.02%), both drops six times daily, with rapid clinical improvement. This treatment was tapered off over 11 months. The keratitis resolved leaving minor, mid-peripheral, anterior stromal scars. She confirmed that she completely discontinued contact lens wear following resolution of the keratitis, but maintained a BSCVA of 20/30.

Five years later, she represented with a 5 month history of increasing discomfort and photophobia in the left eye although there was no associated redness. She had treated these symptoms herself intermittently with non-prescribed, across-the-counter Brolene 0.1% eye drops but as her symptoms failed to settle, she consulted her optometrist who prescribed topical corticosteroids without further investigation. Her symptoms initially improved, however, 3 weeks later she presented to the ophthalmology service with increasing pain and reduced vision. BSCVA was 20/30 OS and slit-lamp examination revealed localized, inferior corneal stromal edema [Figure 2a]. IVCM confirmed Acanthamoeba cysts at the level of the basal epithelium [Figure 2b]. Although there were no obvious cysts visible in the stroma, the view was partially obscured by stromal haze [Figure 2c]. In view of the history, clinical findings, IVCM results, and absence of an epithelial defect; a corneal scrape was not performed. Treatment with topical polyhexamethylene biguanide (PHMB) 0.02% and Brolene 0.1% were commenced hourly by day and the keratitis subsequently improved over the following 4 weeks. The treatment was tapered over 9 months and she retained a BSCVA of 20/30.

Figure 2

(a) Slit-lamp image of the left cornea five years after initial presentation, demonstrating inferior corneal stromal oedema. In vivo confocal microscopy image taken using the Rostock Corneal module showing multiple acanthamoeba cysts at the level of the basal corneal epithelium (b) and stromal haze at 200μm depth (c)

Discussion

To our knowledge, this is the first report of presumed markedly delayed (5 years) recurrence of AK. Although there is a published case of sequential AK caused by genotypically different species of acanthamoeba,[5] the current case is unlikely to represent a new, unrelated, Acanthamoeba infection in the absence of any recent history of subsequent contact lens wear or ocular trauma. Cases of AK without a history of contact lens wear or trauma are rare, with a recent New Zealand study reporting 96% of AK cases are associated with contact lens wear and 4% are associated with trauma.[2] Indeed, the annual incidence of AK is approximately two cases per million in the New Zealand population,[2] indicating a low probability that the second infection was new and unrelated.

A more likely scenario is that despite early clinical resolution and 11 months of intensive treatment with an appropriate combination of anti-acanthamoebal medications, viable cysts persisted and remained dormant in the cornea. These were probably reactivated, although there was no obvious precipitating cause, and the infection was subsequently exacerbated by the introduction of topical corticosteroids.

One possible explanation for the persistence of Acanthamoeba in the reported case is organism resistance to the initial chlorhexidine/propamidine treatment. However, in vitro studies have demonstrated excellent cysticidal performance of chlorhexidine with no demonstrable resistance.[6] Nonetheless, there may be poor correlation between the clinical outcomes of individual cases and in vitro sensitivity,[6] with only 37.5% of patients in one report achieving a medical cure after a prolonged course of treatment.[7] The reasons for the poor in vivo response in the face of good in vitro sensitivity are not fully understood: It is possible that intrastromal therapeutic concentrations are not achieved, that drugs might bind to tissue components, that agents may be inactivated in vivo, that organisms in vivo are more resistant than in vitro, or that acquired drug resistance may occur during the prolonged treatment period.[7]

In vitro studies have clearly demonstrated that the use of topical corticosteroids, even at a minimal dose, promotes excystment of Acanthamoeba cysts and accelerated multiplication of trophozoites.[8] This reflects the role of the introduction of topical steroid in exacerbation of the disease process in the reported case.

The current report is unique in that despite initially mild AK, that was diagnosed at an early stage and treated appropriately over a lengthy period, viable Acanthamoeba cysts remained in the subject's cornea 5 years after the initial episode.[9]

Although the initial episode of AK was proven by culture, there was no microbiological proof of the diagnosis in the recurrent episode. A corneal scrape had not been performed in view of the absence of an epithelial defect. Additionally, the history, clinical findings, IVCM observations, and response to treatment were consistent with a diagnosis of recurrent AK. When performed and interpreted by experienced operators, IVCM has been demonstrated to be highly sensitive and specific in the diagnosis of AK.[10] The authors are therefore confident of the diagnosis of recurrent AK.

This case highlights the importance of caution when using corticosteroids in patients with a previous history of AK. Patients with AK should also be warned regarding the risks of recurrence following presumed resolution.