A Third Candidemia in Eight Months, and This Time the Echinocandin Isn't Working
A single patient, on his third recurrence, this time genuinely resistant rather than simply recurring. The disagreement is less about which drug to use than about correcting an assumption before it becomes the next mistake.
P.R., a 57-year-old man, has been TPN-dependent for over a decade, since a Crohn's-related bowel resection left him with short-gut syndrome and no realistic path back to full oral nutrition. His baseline health, in every respect apart from that dependence, has been stable — he manages his own line care meticulously, by his own account out of necessity rather than any particular fondness for the routine. This is his third episode of candidemia in eight months, each previously treated with micafungin, and each time cultures had cleared and he'd been sent home believing the problem behind him. This time, repeat blood cultures at 72 hours on micafungin remained positive, and susceptibility testing on the current Candida glabrata isolate confirmed reduced echinocandin susceptibility, with sequencing identifying an FKS1 mutation — genuine target-site resistance, not simply another recurrence of a previously susceptible organism.
Salvage options against a confirmed FKS-mutant glabrata isolate are genuinely limited. Liposomal amphotericin B remains the most reliable choice — glabrata usually retains amphotericin B susceptibility even when echinocandin-resistant, and Pappas's IDSA candidiasis guideline names lipid-formulation amphotericin B as the recommended agent where both azole and echinocandin resistance are suspected — but a decade of TPN dependence has left him with real, cumulative renal stress, including documented nephrocalcinosis on prior imaging, that makes standard-dose amphotericin B a genuine cost rather than a free fallback. High-dose fluconazole is an unreliable answer on its own: glabrata already carries intrinsically reduced azole susceptibility even without an FKS mutation, and the efflux and target-modification mechanisms that drive azole resistance can co-occur with echinocandin resistance in the same isolate. What might look like the obvious modern substitute — rezafungin, a newer agent in the same drug class — carries a specific problem worth naming before it's reached for. Perlin's work established FKS-mediated resistance as a mutation in the glucan synthase target itself — the hotspot region every echinocandin binds — rather than a pharmacokinetic limitation specific to the older drugs, which means a longer half-life or less frequent dosing offers no real advantage against a target the whole class can no longer bind effectively. His two prior micafungin courses are the more uncomfortable half of that literature: prior echinocandin exposure is the single best-documented risk factor for acquiring exactly this mutation, which makes his own treatment history part of the explanation for why he is here.
On the consult service, looking at a resistance mechanism, not just a drug name
This is his third episode, and this time we have a confirmed FKS1 mutation, not just another recurrence. Liposomal amphotericin B remains the most reliable option against genuine echinocandin resistance in glabrata — I want to start it today rather than reach for another agent in the same class he's already failed twice.
I agree amphotericin B is the right direction, but a decade of TPN has left real, cumulative renal stress — he has documented nephrocalcinosis on imaging, not just a theoretical risk. I want a cautiously titrated dose with baseline electrolytes and twice-weekly monitoring, coordinated with his TPN team on hydration, not the standard full dose started without that plan in place.
Worth walking through explicitly before we close this out, since rezafungin might otherwise look like the obvious next-generation substitute: on Perlin's account of the mechanism, FKS-mediated resistance is a mutation in the glucan synthase target itself, in the hotspot region shared across the entire echinocandin class. Rezafungin's longer half-life and weekly dosing are real pharmacokinetic advantages elsewhere, but they don't change what the drug binds — against this specific mutation, it offers no meaningful advantage over the micafungin that already failed him. Amphotericin B, titrated the way you're describing, is the actual right call here.
Agreed: liposomal amphotericin B started at a cautiously titrated dose, with baseline renal function and electrolytes, twice-weekly monitoring, and a hydration protocol coordinated directly with his TPN team; all three specialties aligned on this plan once the rezafungin question was walked through explicitly.
Not resolved, and deliberately deferred rather than decided today: whether his central line itself can ever realistically be replaced, or whether recurrent candidemia has simply become the ongoing cost of his TPN dependence. The team agreed this harder conversation belongs at his next outpatient visit, with his long-term care team present, rather than being settled during an acute admission.