Pythium keratitis (Pythium insidiosum keratitis: PIK) is a rare and severe corneal infection caused by the aquatic oomycete Pythium insidiosum. Taxonomically, Pythium belongs to the kingdom Stramenopila, phylum Oomycota, class Oomycetes, order Pythiales, family Pythiaceae, and is not a fungus 1). However, because it closely resembles fungal keratitis clinically, morphologically, and microbiologically, it is also called “parafungus” 1).
Systemic pythiosis was first reported in 1884 by a British veterinarian. The first case of ocular pythiosis was reported from Thailand in 1988, and the first case as a corneal ulcer was reported by Virgile et al. in 1993 in a 31-year-old woman 2). As of 2021, 168 cases of ocular pythiosis have been reported worldwide 2).
Pythium is found in tropical, subtropical, and temperate climates, and has been reported from Thailand, India, China, Australia, Israel, and the United States 1)2). In a cohort from South India, the prevalence of PIK was reported as 5.9% (71/1204 cases) 2). Based on the ITS region or cytochrome oxidase II gene, it is classified into three clades by geographic location: Clade I (ATH: USA), Clade II (BTH: Asia, Australia), and Clade III (CTH: Thailand, USA) 1).
In the largest analysis of 114 cases, the mean age was 41 ± 14.3 years, and 43% were male. 40.4% were farmers, 56.6% were housewives or office workers, and no clear predisposing factor was identified in 55.3% 1).
Pythium is a protist classified as an oomycete, not a fungus. It fundamentally differs from fungi in that its cell wall lacks ergosterol and is composed of cellulose and β-glucan. Therefore, antifungal drugs are ineffective. It also differs from fungi in that it reproduces asexually by biflagellate zoospores.
It presents symptoms common to other microbial keratitis1).
The time from onset to consultation ranges from 2 to 60 days.
The clinical findings of Pythium keratitis are similar to fungal keratitis, but there are characteristic findings1)2).
Characteristic Findings
Reticular dot infiltrate: infiltration with a reticular pattern in the subepithelial to superficial stroma. It suggests a high clinical suspicion of Pythium.
Tentacular projections: tentacle-like infiltrates radiating from the main lesion.
Peripheral furrowing/guttering: Groove-like changes with thinning of the peripheral cornea.
Early limbal involvement: Rapid progression to the limbus, with extension to the sclera.
Findings resembling fungal infection
Feathery margins: Infiltrates with indistinct margins resembling filamentous fungal keratitis.
Satellite lesions/multifocal infiltrates: Multiple small infiltrates scattered throughout the cornea.
Retrocorneal plaque/hypopyon: Signs of deep extension seen in severe cases.
Dry corneal surface: Purulent discharge is scarce, unlike bacterial keratitis.
The median infiltrate size was 5.5 × 6 mm (range: 1–10 mm). The median logMAR visual acuity was 2.78 in 92.1% of patients1).
It is the most common species causing human pythiosis. It has two forms: hyphal and biflagellate zoospore1). Human pythiosis has four clinical forms: vascular, ocular, cutaneous/subcutaneous, and disseminated1).
While fungal keratitis is more common in agricultural workers, Pythium keratitis is characteristically also frequently reported in non-agricultural workers such as housewives, IT professionals, and students1)2).
Infection occurs mainly via zoospores found in tropical and subtropical aquatic environments (paddy fields, ponds, river water, contaminated rainwater). Zoospores adhere to damaged areas of the eye, establishing infection. Cases have been reported not only in agricultural workers but also in urban residents such as housewives and IT professionals.
A high clinical suspicion is necessary for diagnosing Pythium keratitis. If the possibility of Pythium is not specifically communicated to the microbiologist, it may not be detected1).
Smear microscopy of corneal scrapings shows thick, sparsely septate or aseptate hyphae. A ribbon-like folded pattern with right-angle branching is characteristic, but differentiation from fungal hyphae is difficult1)2).
The main staining methods are shown below.
| Staining method | Characteristics |
|---|---|
| 10% KOH + CFW | Sensitivity 79.3–96.5%, specificity >93% |
| IKI-H2SO4 | Pythium-specific, specificity 100% |
| Trypan blue | Sensitivity >75%, no special equipment required |
IKI-H2SO4 (iodine-sulfuric acid) staining stains Pythium blue/blue-black but does not stain fungi, making it extremely useful for differentiation 1)2).
Culture on blood agar medium or potato dextrose agar (PDA). At 37°C, it grows as flat, grayish-white to cream-colored colonies 1)2). It does not grow or grows poorly on Sabouraud agar (with chloramphenicol), which is useful for differentiation from fungi 3). Identification of zoospores by leaf incubation method provides a definitive diagnosis 1)2).
PCR is the gold standard 1)2). Targets include the rDNA-ITS region and the cytochrome oxidase II (cox II) gene.
Metagenomic next-generation sequencing (mNGS) is also an effective method 3). Identification of Pythium by MALDI-TOF mass spectrometry has also been reported 4).
Differentiation from fungal keratitis is most important. According to the Japanese guidelines for infectious keratitis, filamentous fungal keratitis is characterized by feathery ulcers with indistinct borders, and pimaricin is the first-line drug for Fusarium species 5). Since Pythium does not respond to this treatment, the possibility of Pythium should be considered in cases of “fungal keratitis” that do not respond to antifungal drugs.
Other differential diagnoses include Acanthamoeba keratitis (sharing radial keratoneuritis), atypical mycobacterial keratitis (sharing dry gray-white infiltrates), and bacterial keratitis1)2).
Although no established standard treatment protocol exists, recent research is clarifying treatment guidelines.
Antifungal drugs are ineffective against Pythium. Conventional antifungals that inhibit ergosterol synthesis do not act on the cell wall of Pythium, which lacks ergosterol1)2)3).
In an in vitro drug susceptibility test by Hu et al., all six antifungal drugs (fluconazole, itraconazole, voriconazole, posaconazole, amphotericin B, and caspofungin) had MICs >32 mg/mL, showing no antifungal activity3).
Currently recommended drug therapy is antibacterial agents.
In a study evaluating treatment outcomes of 69 eyes, 55.1% were cured with medical therapy (median treatment duration 3 months). The remaining 44.9% required therapeutic corneal transplantation. Cases with infiltrate size >6 mm, long disease duration, and involvement of the posterior stroma were less responsive to medical therapy2).
Therapeutic Corneal Transplantation
Indications: Cases unresponsive to medical therapy, cases with impending corneal perforation.
Key points: Perform with a safety margin of at least 1 mm. Excise including the reticular infiltration pattern1)2).
Recurrence rate: 51.8–54.2% with TPK alone. Decreases to 7.1% when combined with adjuvant therapies such as cryotherapy or ethanol application2).
Other Surgical Interventions
Cyanoacrylate adhesive: Tectonic support for corneal melting. Used in combination with a bandage contact lens1).
Cryotherapy: Single freeze-thaw cycle applied to the limbus with a liquid nitrogen probe. Effective for preventing recurrence2).
Ethanol application: Used when infiltration extends beyond the limbus. Combined with multiple rows of cryotherapy2).
In three cases reported from China, all were initially misdiagnosed as fungal keratitis and treated with antifungal agents. Two cases underwent TPK, and all eventually required enucleation. The cause was considered to be delayed diagnosis due to ineffective antifungal therapy, leading to missed optimal treatment timing4).
Conventional antifungal agents (azoles, polyenes, etc.) target ergosterol, but pythium lacks ergosterol in its cell wall. The cell wall is composed of cellulose and β-glucan, so there is no target for antifungals. Instead, protein synthesis inhibitors such as linezolid and azithromycin are effective.
Recent studies show that in mild to moderate cases, about 55% are cured with antibiotic therapy (linezolid, azithromycin). The recurrence rate in severe cases or after TPK is high (51.8%), but can be reduced to 7.1% with adjuvant cryotherapy and ethanol application. Early diagnosis and appropriate drug selection greatly influence prognosis.
Pythium insidiosum releases biflagellate zoospores (9–10 μm in diameter) from sporangia in aquatic environments1). Zoospores exhibit chemotaxis toward damaged tissue and are attracted to low CO₂ environments1).
The course of infection is as follows:
Zoospores are formed within 1 hour of induction and rapidly produce germ tubes within 24 hours after encystment. This rapid cycle is responsible for the fulminant course of the disease1).
The cell wall of Pythium is composed of cellulose and β-glucan and lacks ergosterol. This characteristic is the underlying cause of antifungal drug resistance, and development of new drugs targeting the cellulose synthesis pathway is expected1)2).
Histopathologically, diffuse destruction of the corneal stroma and neutrophil-predominant inflammatory infiltration are observed. Granulomatous inflammation has been reported in 15% of cases1).
Hu et al. successfully achieved rapid identification of Pythium using metagenomic next-generation sequencing (mNGS). While conventional culture requires 5–7 days, mNGS can rapidly identify even unknown pathogens. However, due to cost issues, it has not yet become widespread as a routine test 3).
Hou et al. created a reference spectrum for Pythium using MALDI-TOF mass spectrometry and added it to their Bruker database, thereby accelerating subsequent Pythium identification 4).
The LAMP method (loop-mediated isothermal amplification) has been reported to have 100% sensitivity and 98% specificity. Since it does not require special equipment, it is expected to be applied in resource-limited facilities 2). Real-time PCR (targeting the exo-1,3-β-glucanase gene) has been reported to have 100% sensitivity and specificity, with a turnaround time of 7.5 hours 2).
Agarwal et al. evaluated the fungicidal effect of ethanol against Pythium in vitro and confirmed its penetration into the cornea using infrared spectroscopy. They reported that the cell membrane of Pythium, which lacks ergosterol, is vulnerable to ethanol, making it a potential treatment option; however, establishing the optimal concentration remains a challenge 2).
Currently, there is no CLSI standard drug susceptibility testing protocol for Pythium. Broth microdilution and disk diffusion methods using zoospores as inoculum have been reported, but variability in methods and inoculum concentrations leads to discrepancies in MIC values 3). Establishing standardized susceptibility testing methods and correlating them with clinical outcomes are future challenges.
