Introduction

A 22-year-old female patient presented with pain and reduced vision in her right eye for the past few days. She had a three-year history of monthly soft contact lens wear and a past medical history of anxiety managed with citalopram. She was otherwise well. The patient worked as a veterinary student, reported to showering in her contact lenses and denied a history of trauma. On presentation right CDVA was 20/120 with spectacles (pinhole 20/30) and 20/16 on the left. The conjunctiva was injected on the right. In the mid peripheral cornea there was a 4 mm by 4 mm epithelial defect within which there was a dense infiltrate measuring 2.2 mm by 1.7 mm. The anterior chamber was quiet. A diagnosis of microbial contact lens related keratitis was made.

A corneal scrape was performed, and levofloxacin 0.3% drops were started hourly day and night. On review day three, right CDVA had improved to 20/80 (pinhole 20/30) and the eye felt more comfortable. However, the infiltrate and epithelial defect had increased in size. Cefuroxime 5% drops were added hourly day and night. On review day 5, the visual acuity had improved to 20/40 on the right with glasses (pinhole 20/30). The right eye remained intermittently sore. On examination, the epithelial defect had increased in size to 8 mm by 8 mm, with the mid peripheral infiltrate now chalky in appearance measuring 2.5 mm by 2.5 mm. Satellite lesions were also apparent. A clinical diagnosis of fungal keratitis was made, which was confirmed by a positive culture of scanty growth of fungus from the initial corneal scrape.

The patient was admitted for treatment with voriconazole 1% hourly day and night, along with cefuroxime and levofloxacin two hourly. Comfort improved on this treatment and the voriconazole was reduced to hourly day only. The patient was discharged on this treatment at day 7. On day 12, the right eye started to deteriorate with increased pain. Visual acuity was reduced to 20/200 on the right (pinhole 20/120). Clinical examination revealed a large endothelial plaque behind the infiltrate that had now enlarged to 3 mm vertical by 2.5 mm horizontal. A further corneal scrape was performed. This and the original scrape identified Metarhizium species.

Treatment was changed to topical natamycin 5% hourly day and night. The corneal ulcer and infiltrate improved rapidly on this treatment and it was tapered to two hourly by day 18. The infiltrate had cleared by day 22. Treatment was reduced to six times a day and continued with the levofloxacin four times a day. At review day 29, visual acuity was 20/80 on the right (pinhole 20/40) with full resolution of the epithelial defect. The right eye was however mildly injected and keratic precipitates were noted on the endothelium. Topical cyclosporin A 0.1% was commenced twice daily with natamycin continued six times daily.

At review day 60, right CDVA was 20/20 with glasses and the eye was comfortable. Clinically there was a 3.5 mm by 3.5 mm scar, not affecting the visual axis. The eye was white, with no epithelial defect or infiltrate and treatment was ceased. The patient was reviewed on day 75 and remained asymptomatic off all treatment with a CDVA of 20/20 and was discharged.

Discussion

Metarhizium anispliae is a filamentous entomopathogenic fungus used worldwide as an agricultural insecticide [1]. The organism has optimal growth at 25 °C and as such has been considered safe, though has extremely rarely being implicated in human disease [1, 2]. Metarhizium has been implicated in a small number of cases of keratitis, sclerokeratitis and endophthalmitis. This report to our knowledge is the ninth worldwide for an ocular infection. The first report was in 1997 from Colombia [3], then subsequent reports emerged from the United States [4, 5], Australia [6], Japan [7], France [2, 8] and Canada [9]. This is the first report in the United Kingdom. Of the previous eight cases, four implicated contact lens wear [4, 5, 7, 9], three involved ocular injection of organic material [2, 6, 8] whilst the initially described case had no identifiable risk factors [3].

Corneal scrape samples were cultured, and fungal elements identified at the North Bristol Bacteriology laboratory. Isolates were sent to the Public Health England South West Mycology Reference Laboratory for identification. The organism was identified as Metarhizium species. Macroscopic and microscopic images are shown in Figs. 1 and 2. Susceptibility studies were performed by CLSI (Clinical Laboratory Standards Institute) broth microdilution method M38-A2. Results are presented in Table 1 with those from the only other two ocular cases with reported susceptibilities [2, 7].

Fig. 1
figure 1

Macroscopic appearance of Metarhizium species identified

Full size image
Fig. 2
figure 2

Microscopic appearance of Metarhizium species identified

Full size image
Table 1 Antifungal susceptibility reports from three Metarhizium keratitis case reports. MIC presented in mg/L
Full size table

Amphotericin-B resistance has been almost universal in ocular and non-ocular Metarhizium mycosis susceptibility reports, indicating a possible intrinsic resistance to the agent [2]. Despite in vitro sensitivity to voriconazole in the reported susceptibilities, there was little response in our case with clinical deterioration. A similar trend was seen with progression despite voriconazole therapy delivered topically [9] or systemically [2] in other cases. It is established that for filamentous fungi particularly, in vitro MIC values indicating susceptibility do not always correlate with effective host response to treatment [10].

Natamycin was used to successfully treat three previous cases with excellent outcomes [3,4,5]. Conversely, two cases progressed to keratoplasty and poor outcome despite natamycin treatment [6, 7]. In the two cases where natamycin was unavailable, infection progressed to require therapeutic keratoplasty despite topical and systemic antifungals, one case with corneal perforation at day 11 of treatment [2, 9].

Natamycin 5% is employed as a first line empiric antimycotic therapy in many centres [10]. Potential drawbacks include limited ocular penetration, lack of widespread availability of a commercial ophthalmic product and restricted efficacy in yeast type keratomycoses. Empiric first-line treatment with natamycin 5% is optimal for superficial mycotic appearing infiltrates or those with filamentous elements identified on microscopy or culture [10]. In the context of Metarhizium species identification, this case adds support to the use of topical natamycin. This report further demonstrates clear clinical inferiority of voriconazole despite in vitro sensitivity as well as the presence of amphotericin B resistance in Metarhizium keratitis samples.