Candidemia Cleared From the Blood, Not Yet From the Eye
A single patient, cleared systemically and about to step down early. The disagreement is whether an eye exam most cases never need has just changed what "controlled" actually means.
D.O., a 58-year-old man, was three weeks into a slow recovery from necrotizing pancreatitis — his second hospitalization for it, both tied to a longstanding struggle with alcohol he has finally started addressing with an outpatient program he was due to begin the week he got readmitted — when a fever spike on hospital day 19 turned into blood cultures growing Candida albicans, sourced most likely to the central line placed for his TPN. Micafungin was started immediately, the line pulled two days later once a second access site was secured, and by hospital day 23 his repeat cultures had cleared. He was already being scheduled for step-down to oral fluconazole, the standard move once a fluconazole-susceptible isolate clears from the blood.
The dilated eye exam ordered per protocol for every case of candidemia — before that step-down could happen — found something the blood cultures alone would have missed: chorioretinal lesions consistent with early Candida chorioretinitis, no vitreous involvement yet. That finding reframes what looked like a routine transition. Echinocandins like micafungin achieve only minimal vitreous concentrations regardless of how well they've cleared the bloodstream, which means the drug currently controlling his systemic infection is doing essentially nothing for the one organ now shown to be involved — while fluconazole, and voriconazole after it, both cross the blood-ocular barrier well enough to be the actual treatment for ocular disease. The IDSA candidiasis guideline treats chorioretinitis without vitritis as needing a minimum of four to six weeks of therapy with an agent that reaches the eye, often at a higher fluconazole dose than the standard candidemia course — a genuinely different regimen than the one he was about to be sent home on, and a fourteen-day course he had already mentally finished would in reality only just be starting.
At the bedside, after the eye exam changes the plan
Micafungin has essentially no vitreous penetration, whatever his blood cultures show. Right now the drug treating his eye is doing nothing, and this isn't a case where we can afford to finish out a standard step-down timeline before switching — I want fluconazole started today, at the dose the ocular literature actually supports, not the routine candidemia dose.
You're right that echinocandins don't reach the eye — I'm not arguing that. But the line came out two days ago, and I want two things confirmed before I call this systemically controlled, not zero: sustained negative cultures and a clean recheck of the new access site. That's a same-day answer, not a delay measured in weeks.
Switching him off micafungin before source control is fully confirmed treats the eye finding as the only open question, when the line itself was only pulled 48 hours ago.
Both points can be satisfied on the same timeline. Once today's cultures confirm clearance, start fluconazole at 800mg daily — double the routine step-down dose — specifically because chorioretinitis needs higher ocular concentrations than bloodstream clearance alone requires, for a minimum six-week course with an ophthalmology recheck at two weeks to confirm the lesions are actually regressing and not just being treated on paper.
Agreed: same-day blood cultures confirmed clear, high-dose fluconazole started that afternoon, and a two-week ophthalmology recheck scheduled to confirm the lesions are regressing rather than assuming the switch alone settles the question.
Fully resolved — no disagreement carried forward. The ID physician's source-control concern and the ophthalmologist's urgency were both genuinely satisfied by the same-day culture result, and all three specialties signed off on the combined plan before he left the unit.