Key messages

• The use of natamycin in fungal keratitis compared with voriconazole is probably associated with better visual acuity.
• The use of natamycin in fungal keratitis compared to voriconazole prevents corneal perforation and/or need for therapeutic keratoplasty.
• The balance is probably favorable to the use of natamycin in fungal keratitis.

What is the evidence.
See evidence matrix in Epistemonikos later

We found three systematic reviews [1], [2], [3] that included three primary studies reported in seven references [4], [5], [6], [7], [8], [9], [10], all corresponding to randomized controlled trials.

What types of patients were included*

All trials included male and female patients older than 16 years with an average age of 46.7 years, with microbiological evidence of fungal keratitis; direct visualization of hyphae in corneal scraping by smear with 10% KOH stain or Gram stain [5], smear with 10% KOH stain, Giemsa stain or Gram stain [4] and positive corneal ulcer smear for fungi in patients with visual acuity of 20/40 (0.3 logMAR units) at 20/400 (1.3 logMAR units) [8].

All trials excluded patients who had concomitant corneal coinfection (viral, bacterial or protozoal), signs of imminent corneal perforation or history of corneal perforation [4], [5], [8]. Two trials [4], [8] excluded patients with visual acuity worse than 20/200 (1.0 logMAR units) in the unaffected eye and with bilateral ulcers. Two trials excluded patients without light perception in the affected eye [4], [5], and one trial also excluded pregnant patients [4].

What types of interventions were included*

All trials used topical 5% natamycin as intervention: two trials administered 1 drop every hour for 1 week and then 1 drop every 2 hours for 2 weeks (while the patient was awake) [4], [8] and one trial 1 drop every hour for 2 weeks [5].

As comparison, topical 1% voriconazole was used: in two trials 1 drop was administered every hour for 1 week and then every 2 hours for 2 weeks (while the patient was awake) [4], [8] and one trial administered 1 drop each hour for 2 weeks [5].

What types of outcomes
were measured

The trials evaluated multiple outcomes, which were grouped by the systematic reviews as follows:

• Clinical cure (at 2-3 months)
• Time to achieve clinical cure
• Visual acuity (at 2-3 months)
• Complications of fungal keratitis (corneal perforation, therapeutic keratoplasty, enucleation)
• Treatment failure
• Adverse effects, including: corneal thinning or descemetocele formation, corneal perforation, endophthalmitis, chemosis, punctate keratopathy, recurrent epithelial erosions, conjunctival injection, ulceration and necrosis of conjunctiva, hepatic and renal toxicity)
• Quality of life

The average follow-up of the trials was 11.3 weeks, with a range between 10 and 12 weeks.

To whom this evidence does and does not apply

• The evidence presented in this summary apply to adult patients with filamentous fungal keratitis (Fusarium, Aspergillus and Curvularia genus).
• Although the identified trials evaluated only adults, and no evidence was found in adolescent and pediatric patients with filamentous fungal keratitis, it is reasonable to extrapolate these results to those populations, in absence of direct evidence.
• These results do not apply to patients with fungal keratitis due to yeast, such as Candida spp.
• These results also do not apply to patients with infectious keratitis of bacterial, viral, parasitic, or mixed etiology (for example coinfection due to fungi and bacteria), since the treatment of infectious keratitis depends on its specific cause.

• The outcomes analyzed in the summary of findings table are those considered critical for decision-making by the authors of this summary. In general, they coincide with those evaluated in the systematic reviews identified.
• The clinical cure outcome was selected because it directly allows to determine the efficacy of the treatment for fungal keratitis.
• Visual acuity is the outcome with the greatest clinical importance; it is correlated with severity or improvement of the disease and it is in general the visual result that is most relevant for the patients, since their compromise directly affects the overall functioning and quality of life.
• Corneal perforation and/or need to perform therapeutic keratoplasty was selected as the critical outcome in the course of the disease, since its appearance leads to a more severe disease with a high risk of developing fungal endophthalmitis, and in consequence, to the need to establish a more aggressive treatment (including therapeutic keratoplasty).

• The use of topical 5% natamycin in fungal keratitis, compared with topical 1% voriconazole, could result in a higher probability of clinical cure, but the existing evidence has certain limitations. However, the use of topical 5% natamycin prevents corneal perforation and/or need to perform therapeutic keratoplasty and is probably associated with better visual acuity compared to topical 1% voriconazole, with high and moderate certainty of the evidence respectively.
• Regarding the adverse effects of the treatments, the systematic reviews [1], [2] indicated that there were no adverse side effects reported in the three randomized trials evaluated [4], [5], [8]. The literature reports that topical 5% natamycin is well tolerated at the corneal level, however its prolonged use presents toxicity generally in the form of punctate keratitis. On the other hand, topical 1% voriconazole is well tolerated, with no reports of side effects to its use [11].
• The use of topical 5% natamycin in fungal keratitis has greater benefits compared to the use of topical 1% voriconazole, with an equivalent safety profile between both treatments. Therefore, we consider that there is probably a favorable balance between benefits and risks for natamycin over voriconazole.

• Both natamycin and voriconazole are generally high cost treatments, but the treatment of fungal keratitis with natamycin is about half the value associated to the treatment with voriconazole [13]. In addition, considering that the use of natamycin in fungal keratitis compared with voriconazole decreases the number of patients that evolve with corneal perforation and/or need to perform therapeutic keratoplasty, it is likely that the use of natamycin in fungal keratitis is cost-effective compared to the use of voriconazole.
• It is important to consider that in health systems with limited resources, it is possible that none of the evaluated alternatives is available, so alternative treatments of lower economic cost and greater availability could be considered, for example amphotericin or chlorhexidine. The comparative efficacy between the latter and natamycin or voriconazole was not evaluated in the present summary.

• Faced with the evidence presented in this article, most patients and clinicians should lean in favor of the use of natamycin over voriconazole in fungal keratitis.
• It is important to note that in a context of low resources, probably due to the high cost of natamycin, patients may prefer lower cost treatments, such as amphotericin or chlorhexidine.

Differences between this summary and other sources

• The conclusions of this summary agree with those of the systematic reviews analyzed.
• The guideline for the management of corneal ulcer of the The World Health Organization developed by the Regional Office for South-East Asia recommends topical 5% natamycin or topical 0.15% amphotericin as the first-line treatment for filamentous fungal keratitis [14]. The American Academy of Ophthalmology (AAO) recommends topical 5% natamycin as the treatment of choice for filamentous fungal keratitis, with the exception of deep or severe infections (due to the low penetration of natamycin), where it is recommended the joint use of topical 5% natamycin with topical 1% voriconazole [15]. The National Health Service (NHS) of the United Kingdom in its guideline about infectious keratitis, recommends topical 5% natamycin as a first line treatment and as a second line topical 0.02% chlorhexidine, topical 1% voriconazole or topical 5% amphotericin [16].

• The probability that future research changes the conclusions of this summary regarding the clinical cure and change in visual acuity is high, due to the existing uncertainty. It is unlikely that will change the conclusions about corneal perforation and/or need to perform therapeutic keratoplasty, since its certainty of the evidence is high.
• We found one randomized trial not included in the systematic reviews identified, comparing topical 5% natamycin versus topical 1% voriconazole [13]. 
• We did not identify ongoing trials addressing this question in the International Clinical Trials Registry Platform of the World Health Organization.We did not identify ongoing systematic reviews in the International Prospective Register of Systematic Reviews (PROSPERO) of the National Institute for Health Research.