Optic-Nerve and Corneal Drug Toxicity

Two targets, two very different stories

The optic nerve and the cornea sit at opposite ends of the eye — and of the danger scale. The optic nerve is the eye's output cable: roughly a million retinal ganglion-cell axons carrying the finished image to the brain. Being neural tissue, it does not regenerate, so damage here threatens permanent, sight-losing injury. The cornea is the eye's transparent front window, and its clarity depends on a smooth epithelium and a precisely ordered stroma; certain drugs deposit in it and cloud that clarity, but the cornea's surface renews itself constantly, so most corneal drug effects are reversible and often symptomless. That contrast frames the whole chapter: optic-nerve toxicity is the one you must never miss, while corneal deposits are usually a harmless marker that a drug is on board. Both belong to the same map as the retinal-toxicity chapter — the shared cross-section of the eye there tags the optic nerve (ethambutol, amiodarone) and the cornea (amiodarone/chloroquine verticillata) as two of the four classic sites of systemic-drug injury.

Toxic optic neuropathy: the pattern to memorise

Four words carry the diagnosis: painless, bilateral, symmetrical, central. Toxic optic neuropathy has a signature that separates it from almost everything else that hits the optic nerve. It is painless — unlike the optic neuritis of multiple sclerosis, which typically aches on eye movement. It is bilateral and roughly symmetrical, because a circulating drug reaches both nerves equally — unlike a stroke of the optic nerve, which strikes one eye. And it is central: it attacks the papillomacular bundle, the delicate fibres serving fine central and colour vision, so patients lose central acuity and colour discrimination (dyschromatopsia) while their peripheral field is spared. The earliest and most sensitive sign is often a loss of colour vision — reds desaturate first — which is exactly why formal colour-vision testing is the workhorse of monitoring. Because the injury is metabolic rather than structural, catching it early and withdrawing the drug usually allows recovery; leaving it too long lets the axons die and the optic disc turn pale, at which point the loss is permanent.

Ethambutol: the classic optic-nerve toxin

The one drug on this page every student must know cold — and the reason TB clinics do baseline eye checks. Ethambutol is a first-line anti-tuberculous drug — the "E" of the standard RIPE regimen taught in the Antimicrobials chapters — and its dose-limiting toxicity is optic neuropathy. The risk is dose- and duration-dependent: it climbs with higher milligram-per-kilogram doses and with longer courses, and it is worse in renal impairment, because the kidney clears the drug and a failing kidney lets levels build. The proposed mechanism is a metabolic insult — ethambutol chelates metals and disrupts mitochondrial function in the energy-hungry ganglion-cell axons. The clinical picture is the textbook one: painless, bilateral loss of central acuity and colour vision, classically red-green dyschromatopsia. The saving grace is that it is usually reversible if the drug is stopped promptly — which is why patients are counselled to report any visual change at once, and why a baseline acuity and colour-vision assessment is done before starting, with regular monitoring during treatment. Isoniazid, another drug in the same regimen, can rarely cause its own optic neuritis and adds to the picture.

The wider gallery of optic-nerve toxins

Beyond ethambutol, a handful of agents produce the same painless bilateral picture. Amiodarone, the antiarrhythmic, causes an insidious optic neuropathy that is distinct from — and must not be confused with — the corneal deposits it also produces; the neuropathy is a genuine threat to vision, the deposits usually harmless, and one patient can have both. Linezolid, an oxazolidinone antibiotic, causes optic (and peripheral) neuropathy with prolonged use beyond the usual short course — a mitochondrial toxicity that limits long treatments and is covered again in the Antimicrobials material. Isoniazid, as noted, can inflame the optic nerve, a risk linked to pyridoxine (vitamin B6) depletion, which is why B6 is co-prescribed with it. Infliximab and the other anti-TNF biologics can rarely trigger optic neuritis and demyelination. Sildenafil and the other PDE5 inhibitors carry a rare association with non-arteritic anterior ischaemic optic neuropathy (NAION) — a sudden, painless, usually one-eye event distinct from the slow toxic pattern. And methanol, from the Toxicology chapter, is the devastating outlier: its metabolite formic acid poisons the optic nerve directly, causing bilateral blindness — the reason methanol poisoning is an emergency treated with fomepizole or ethanol.

Corneal deposits: vortex keratopathy (verticillata)

A whorl of golden-brown lines swept across the cornea like iron filings around a magnet. The signature corneal drug effect is cornea verticillata — vortex keratopathy — a whorl-like pattern of fine greyish or golden-brown deposits in the corneal epithelium, spiralling out from a point just below the centre. The mechanism unites a whole drug class: cationic amphiphilic drugs bind cellular lipids and accumulate as lysosomal deposits inside the epithelial cells, which then migrate in the cornea's natural swirl pattern. The most common culprit by far is amiodarone — verticillata appears in the great majority of patients on long-term therapy — but the same whorls are caused by chloroquine and hydroxychloroquine (the very drugs whose retinal toxicity fills the sibling chapter), by tamoxifen, and by indomethacin, among others. The reassuring point is that vision is almost always spared: patients may occasionally notice haloes around lights, but acuity is typically normal, and the deposits are reversible, fading over months once the drug is stopped. So verticillata is rarely a reason to change treatment; its real value is as a clue — a visible fingerprint that a patient is taking, or has been taking, one of these drugs.

The wider anterior-segment gallery

Beyond deposits, several drugs injure the ocular surface or trip the front of the eye in ways worth a warning. Two forms of surface toxicity come from the treatment itself. Chronic topical anaesthetic abuse is the notorious one — repeated self-administration of drops like proxymetacaine or tetracaine (often by a desperate patient, or a healthcare worker with access) is directly toxic to the corneal epithelium and can melt the cornea into a persistent epithelial defect and a ring-shaped ulcer; this is why topical anaesthetics are never dispensed for home use, a rule reinforced in the Foundations and anaesthetics chapters. Quieter but far more common is preservative toxicity: benzalkonium chloride (BAK), the preservative in most multi-dose eye drops, is a detergent that damages the ocular surface with long-term use, producing a punctate keratopathy and dry, sore eyes — the reason preservative-free formulations exist for patients on lifelong drops such as glaucoma therapy. Then come three systemic-drug traps for the front of the eye: tamsulosin and the other alpha-blockers cause intraoperative floppy iris syndrome (IFIS) — a flaccid, billowing iris that can persist even after the drug is stopped, so the cataract surgeon must be warned in advance. Bisphosphonates, especially the intravenous ones, can trigger acute anterior uveitis and scleritis. And topiramate and the sulfonamides can precipitate acute bilateral angle-closure glaucoma through a ciliochoroidal effusion that shoves the lens-iris diaphragm forward — an idiosyncratic mechanism quite unlike ordinary pupillary-block angle closure, so the drug must be stopped and pilocarpine is not the answer. That contrast connects to the angle-closure and mydriatics chapters, where the pupillary-block story is told in full.

The unifying lessons

Stand back and three principles organise the whole chapter. First, sort by danger: painless bilateral central visual loss with dyschromatopsia is toxic optic neuropathy — the sight-threatening emergency that demands you stop the drug — whereas corneal whorls are, in themselves, a benign marker of exposure that rarely changes management. Never let the harmless finding distract you from the dangerous one, especially with amiodarone, which can cause both. Second, some of these drugs come with a formal monitoring contract, exactly as hydroxychloroquine did in the retinal chapter: ethambutol demands baseline and regular colour-vision and acuity checks, and the responsibility to counsel the patient to report any visual change at once. Third, know your reversibility: most toxic optic neuropathy and essentially all verticillata recover if the drug is withdrawn in time, but a neglected optic neuropathy leaves a pale disc and permanent loss. As always, the single most useful test in unexplained visual change is a careful drug history — the ethambutol on a TB chart, or the amiodarone in a cardiac one, may be the whole diagnosis.