Post-traumatic fungal keratitis after a wheat stalk injury
A 46-year-old wheat farmer presents with a five-day history of a painful right eye after being struck by a wheat stalk while harvesting. A rural clinic has treated him with topical tobramycin four times daily and dexamethasone twice daily since day two. Visual acuity is 6/60 in the right eye and 6/9 in the left. There is a 4 × 4 mm paracentral stromal infiltrate with feathery margins, several smaller satellite infiltrates, a 1.5 mm hypopyon and moderate anterior chamber cells; the intraocular pressure is 12 mmHg and the pupil is round. A corneal scrape stained with potassium hydroxide shows septate branching hyphae; Gram stain shows no organisms. He lives 200 km from the hospital, has no access to natamycin locally, and asks whether the drops he already has will be enough.
1. What is your immediate management, including the steroid he is already using?
Read the worked answer
The history (organic matter, agricultural work), the morphology (feathery margins, satellite lesions, hypopyon with a relatively quiet anterior chamber relative to the size of the infiltrate) and the potassium hydroxide finding of septate branching hyphae make this filamentous fungal keratitis. Two things he is using are wrong: the steroid and the inadequate antibacterial.
Immediate steps:
- Stop the topical dexamethasone. Steroids promote fungal proliferation and increase the risk of perforation. I would stop it outright rather than taper, and document why.
- Start intensive topical antifungal therapy. Natamycin 5% is first-line for filamentous keratitis, hourly day and night initially. If it is genuinely unavailable, topical voriconazole 1% hourly is a reasonable substitute, and systemic therapy (oral voriconazole or itraconazole) should be added for a deep or large ulcer with hypopyon, after checking liver function and with attention to cost and interactions.
- Admit him. He cannot manage hourly drops at 200 km and needs daily slit-lamp monitoring for thinning and perforation.
- Debride the epithelium over the infiltrate to improve drug penetration, and consider a corneal scrape for culture on blood agar, Sabouraud agar and thioglycollate broth even though KOH is positive - speciation and sensitivity guide escalation.
- Cycloplegia for comfort and to prevent synechiae, and analgesia.
On his question about the drops he has: the tobramycin is not treating the fungus and the dexamethasone is actively harmful. I would say this plainly, and explain that fungal ulcers heal over weeks, not days, and that the aim in the first fortnight is to prevent perforation.
2. What would make you add antibacterial cover, and when would you proceed to surgery?
Read the worked answer
Adding antibacterial cover: the Gram stain is negative, but a scrape is only about 50–60% sensitive and mixed bacterial–fungal infection is common after organic injury. I would add a broad-spectrum topical (for example fortified cefazolin plus tobramycin, or a fluoroquinolone) if there is purulent discharge, a rapidly enlarging infiltrate, worsening hypopyon or anterior chamber reaction, or if culture subsequently grows bacteria. Routine dual cover in a purely fungal ulcer risks added toxicity on an already compromised cornea, so I would not add it reflexly - I would decide at each daily review on the clinical trajectory.
Proceeding to surgery - therapeutic penetrating keratoplasty - I would consider if:
- The cornea is perforating or at imminent risk (descemetocele, Seidel-positive, depth of the infiltrate reaching the endothelium beyond the limbus of safe medical therapy).
- The infiltrate continues to enlarge after 1–2 weeks of correctly administered, adequate antifungal therapy, especially with a growing hypopyon or extension to the sclera.
- There is impending or actual intraocular spread.
I would avoid lamellar keratoplasty when the infiltrate is full thickness, since it leaves fungus behind. Before surgery, the eye should have as much antifungal load as possible; tectonic keratoplasty with a large graft and adjunctive systemic antifungal therapy gives the best chance. I would also counsel on the guarded prognosis: fungal ulcers with hypopyon have a high rate of perforation, a real risk of endophthalmitis, and a substantial chance of needing a second graft or of permanent visual loss even with ideal care.