The cornea is the eye's primary refractive surface — responsible for approximately two-thirds of the eye's total refractive power. It is avascular (no blood vessels), deriving nutrition from the tear film anteriorly and aqueous humor posteriorly. This avascularity is essential for optical clarity but creates a unique vulnerability: without the rapid cellular immune response that vascularity enables, the cornea relies almost entirely on its surface defences to repel microbial invasion.

The corneal epithelium is the first line of defence — a 5–6 cell layer that regenerates completely within 72 hours when scraped cleanly. Its tight junctions, mucin glycocalyx (from the tear film), and inherent resistance to most organisms mean that an intact epithelium is sufficient protection against the vast majority of environmental microbes. The moment this layer is breached — by a scratch, a contact lens abrasion, a chemical burn, or a dry eye surface defect — pathogens have direct access to the avascular stroma.

Once in the stroma, bacteria proliferate rapidly, releasing proteases that liquefy collagen and create a "melting" appearance. Fungi grow slowly by sending thread-like hyphae between collagen lamellae — an infiltrative pattern that makes them resistant to penetration by topical drops. Acanthamoeba cysts are extraordinarily resistant to almost all antimicrobials and can lie dormant in the stroma for weeks. The structural consequences — permanent stromal scar, corneal thinning, descemetocele, or perforation — depend entirely on how quickly effective treatment is initiated.

India carries a corneal blindness burden that is disproportionate even relative to its population size. The confluence of factors that create this burden is well understood — and entirely preventable in most cases.

Agricultural trauma is the defining factor. Approximately 60% of India's population depends on agriculture. Harvesting operations — particularly paddy, wheat, sugarcane, and bamboo cutting — produce a steady stream of vegetative eye injuries. Plant material (especially paddy husks and stalks) carries filamentous fungi on its surface. When these penetrate the cornea, they inoculate the stroma directly with Fusarium, Aspergillus, or Curvularia species. Unlike bacterial keratitis, which presents dramatically within 24–48 hours, fungal keratitis has an insidious onset — the patient may not present for 5–10 days, by which time stromal involvement is deep.

Delayed presentation compounds the problem. Rural patients often consult traditional healers first, receiving plant-based eye drops that may introduce additional organisms. Pharmacists dispense topical antibiotics — appropriate for bacterial infections — that have no activity against fungi. By the time these patients reach a tertiary eye hospital, the ulcer may have progressed to involve the entire stroma or reached the posterior surface. A study from L.V. Prasad Eye Institute found that the median time from injury to first ophthalmic consultation was 7.2 days — and from consultation to correct diagnosis, an additional 4.8 days in fungal cases.

The corneal transplant gap makes the downstream consequence stark. India needs approximately 25,000 corneal transplants annually; fewer than 3,000 are performed, owing to a severely underdeveloped eye banking infrastructure and cultural barriers to eye donation. The result: tens of thousands of patients who could have their vision restored by a surgical procedure are permanently monocular or blind.

The most dangerous misconception in keratitis management is treating all red eyes with a corneal opacity as if they are bacterial. The four main pathogen categories — bacterial, fungal, viral, and amoebic — require entirely different treatment approaches, and the wrong empirical treatment does not merely fail to work: it actively worsens outcomes. Steroids given to fungal keratitis allow explosive fungal proliferation. Antifungals given to bacterial keratitis delay appropriate treatment. The only way to avoid this is microbiological diagnosis before initiating specific treatment.

The visual impact of a corneal ulcer depends on its location, depth, and the degree of associated stromal oedema and scarring. A peripheral ulcer may cause minimal visual disturbance; a central infiltrate 1mm in diameter can reduce acuity from 6/6 to counting fingers. The interactive simulation below demonstrates how progressive corneal opacity affects the visual experience.

The single most important clinical rule in infectious keratitis management is this: do not start specific antimicrobial treatment without corneal scraping and smear examination. In practice, broad-spectrum empirical antibiotics are often started while awaiting culture results — but the smear (Gram stain + KOH mount) takes 30 minutes and should be done before the first drop is prescribed. A positive KOH mount showing fungal hyphae changes the entire treatment protocol.

The ulcer is assessed under white light and then cobalt blue filter after fluorescein installation (using a FLUROSCÉNE fluorescein strip moistened with saline). The epithelial defect stains bright green — its size, shape, and edges provide diagnostic clues: branching dendritic pattern (HSV), feathery edges with satellite lesions (fungal), sharp-bordered grey-white infiltrate (bacterial), or ring infiltrate following nerve distribution (Acanthamoeba). Rose bengal staining reveals devitalised epithelial cells not visible with fluorescein — particularly useful for HSV and dry eye-related keratitis. The depth of stromal involvement, presence of hypopyon, and status of the posterior cornea (descemetocele — a bubble of Descemet's membrane protruding through thinned stroma — indicates imminent perforation) are documented.

Under topical anaesthesia (proxymetacaine or benoxinate drops), a sterile Kimura spatula or disposable blade is used to scrape the leading edge and base of the ulcer. Material is spread directly onto glass slides for Gram stain (bacteria) and KOH mount (fungi), and inoculated onto culture media — blood agar (aerobic bacteria), chocolate agar (Haemophilus, Neisseria), Sabouraud dextrose agar (fungi), thioglycolate broth (anaerobes), and non-nutrient agar with E. coli lawn (Acanthamoeba). The scraping is repeated 2–3 times to ensure adequate material from different depths.

KOH mount result time: 30 minutes. If hyphae are seen — treat as fungal. If Gram-positive cocci in clusters — Staphylococcus. Gram-negative rods — Pseudomonas or Klebsiella. Starting treatment before any smear result is returned delays appropriate care by hours in bacterial cases and days to weeks in fungal cases.

IVCM is the most powerful diagnostic tool for infectious keratitis — providing real-time, non-invasive cellular-resolution imaging of the corneal layers. Under IVCM, fungal hyphae appear as highly reflective branching structures within the stroma; Acanthamoeba cysts appear as round, double-walled highly reflective structures 15–30 µm in diameter; corneal nerves can be visualised and their involvement in keratitis assessed. IVCM detects fungal keratitis with sensitivity of 89% and specificity of 93% in experienced hands.

The limitation is availability — IVCM instruments are present in very few centres outside major Indian cities, leaving the majority of keratitis diagnoses dependent on smear and culture. AI-assisted IVCM image analysis is in early clinical trials and may eventually enable remote diagnosis.

Moxifloxacin 0.5% (broad-spectrum fluoroquinolone) is the standard first-line empirical antibiotic for bacterial keratitis — every 30–60 minutes while awake for the first 48 hours, then tapering based on response. For large central ulcers, severe presentation, or suspected Pseudomonas (contact lens-associated, rapid melting), fortified drops — cefazolin 5% + tobramycin 1.3% alternating hourly — provide superior bacterial coverage. Systemic antibiotics add little unless there is impending or actual perforation with risk of endophthalmitis. Response is assessed at 24–48 hours: a responding ulcer shows reduced infiltrate density and smaller epithelial defect.

Natamycin 5% suspension — topically every 1–2 hours, tapered over weeks — is the first-line for filamentous fungal keratitis (Fusarium, Aspergillus). It works by binding to ergosterol in the fungal cell membrane. The major challenge is penetration: natamycin does not penetrate well into deep stroma, and fungal hyphae can extend beyond the visible infiltrate. Voriconazole 1% (compounded eye drops or oral voriconazole in severe cases) is preferred for Candida keratitis and voriconazole-sensitive moulds. Amphotericin B 0.15% is an alternative for Aspergillus. Treatment duration: minimum 6–8 weeks, often 3–4 months for deep infections. Any patient not responding after 2–3 weeks should be referred for surgical evaluation.

The combination of PHMB 0.02% (polyhexamethylene biguanide) + chlorhexidine 0.02% — coapplied every 1–2 hours — is the current standard for Acanthamoeba keratitis. Both agents target trophozoites effectively; neither reliably kills encyst forms at standard concentrations. This is why treatment must continue for months even after apparent clinical resolution, as cysts can re-emerge. Propamidine isethionate (Brolene) 0.1% is added in many centres. Pain management is a significant clinical challenge — the neuropathic corneal pain in Acanthamoeba keratitis is severe and may require oral neuropathic agents. Penetrating keratoplasty — when indicated — should ideally wait until the infection is completely quiescent, as surgical manipulation can reactivate dormant cysts.

Epithelial HSV keratitis (dendritic, geographic) is treated with topical ganciclovir 0.15% gel (5× daily) or acyclovir 3% ointment — avoiding débridement which can spread the virus. Systemic acyclovir or valacyclovir is added for severe or recurrent cases. Stromal keratitis (interstitial, disciform) — which is immune-mediated, not directly viral — is treated with topical corticosteroids + antivirals simultaneously (Herpetic Eye Disease Study, HEDS protocol). This is the one exception to the "no steroids in infectious keratitis" rule — but it requires confirmation of stromal (not epithelial) involvement and antiviral cover. Long-term oral acyclovir prophylaxis (400 mg BD) reduces annual recurrence rate by 45% (HEDS trial).

Therapeutic penetrating keratoplasty (TPK) — a full-thickness corneal transplant performed not to restore vision but to eliminate infection and prevent globe perforation — is required in approximately 10–15% of fungal keratitis cases and a smaller proportion of bacterial cases. The indications are:

• Progressive stromal melting despite maximal medical therapy
• Descemetocele (Descemet's membrane exposed — imminent perforation)
• Corneal perforation with iris prolapse (emergency TPK)
• Deep endophthalmitic extension (infection reaching the anterior chamber)
• No microbiological response after 3–4 weeks of appropriate treatment

The timing is critical and counterintuitive: performing keratoplasty during active infection carries very high graft failure rates due to ongoing inflammation and enzyme-mediated graft melting. Where possible, medical stabilisation for 2–4 weeks before surgery improves outcomes. However, a perforated eye or one close to perforation cannot wait — emergency TPK accepts a higher failure rate as preferable to globe loss.

For eyes that have been rendered blind and painful by a healed corneal scar — the majority of India's 6.8 million corneal blind patients — optical penetrating keratoplasty offers the possibility of visual rehabilitation. The Ramayamma International Eye Bank at L.V. Prasad Eye Institute, the Eye Bank Association of India, and newer initiatives like the National Programme for Control of Blindness (NPCB) corneal transplant programme are working to close the 8× gap between need and availability.

The overwhelming majority of corneal blindness from infectious keratitis is preventable. Unlike retinal diseases where the damage mechanism is metabolic and systemic, corneal ulcers begin with a single preventable event — an unprotected eye injury, a contact lens left in water, a delay in seeking care. Prevention operates at three levels:

• Primary prevention (avoiding injury): Eye protection during agricultural work — safety goggles during harvesting operations — is estimated to prevent up to 40% of fungal keratitis cases. Awareness campaigns targeting rural farmers have been trialled by L.V. Prasad Eye Institute's community outreach programme with measurable reduction in agricultural keratitis cases in enrolled villages. The economics are stark: a pair of safety spectacles costs ₹50–100; a tertiary hospital keratitis admission costs ₹5,000–50,000; a corneal transplant, if available, costs ₹40,000–1,50,000.
• Contact lens hygiene: No sleeping in contact lenses. No showering, swimming, or using tap water with lenses. Lens cases cleaned daily and replaced monthly. Any red eye, pain, or discharge = remove lenses immediately and seek same-day review. Acanthamoeba and Pseudomonas keratitis — both potentially devastating — are almost entirely contact-lens-related in urban India.
• Secondary prevention (early treatment): Any agricultural eye injury should be seen by an ophthalmologist within 24 hours — not a pharmacist, not a traditional healer, not a "wait and see" approach. The delay from injury to ophthalmic consultation is the single most modifiable factor in fungal keratitis outcome.