Candidemia in a Line That Is Also the Only Working Dialysis Access Left
A single patient, whose infected catheter is also her only way to survive the week. The disagreement is whether guideline-recommended removal still applies when the alternative isn't actually available yet.
E.T., a 68-year-old woman, has been on hemodialysis for six years, and her tunneled dialysis catheter is not a convenience — it is what stands between her and the three-times-weekly treatment that keeps her alive, after two AV fistula failures and a central vein stenosis that has quietly closed off most of the usual next-site options. Her baseline health, apart from ESRD itself, has been notably stable; she has kept every dialysis appointment for the past year without a single missed session. She was admitted two days ago with fever and rigors during a dialysis run, and blood cultures drawn from both the catheter and a peripheral site grew Candida parapsilosis in both, consistent with a catheter-associated source. She is hemodynamically stable, on micafungin since culture results returned.
The IDSA candidiasis guideline is direct about catheter removal for candidemia — it strongly recommends it when feasible, and the mortality data behind that recommendation is real, not a formality. But "when feasible" is doing genuine work in her case rather than being a routine caveat: the vascular access surgeon has already stated that a new permanent access site isn't readily available for at least one to two weeks given how much of her usual anatomy is already used up, and a temporary non-tunneled catheter carries its own continuation-of-infection risk rather than solving the problem outright. Micafungin was started empirically before speciation returned, and the species that came back complicates the choice rather than confirming it. C. parapsilosis carries a naturally occurring FKS1 polymorphism that raises echinocandin MICs across the class, which is why the same IDSA guideline recommends fluconazole for confirmed C. parapsilosis infection and singles this species out — alongside C. glabrata — as one where echinocandin susceptibility should actually be tested rather than assumed. That points away from the drug she is on. What points back toward it is the line itself: echinocandins retain meaningfully better in vitro activity against Candida biofilm than fluconazole does, and biofilm is precisely what a retained, colonized catheter is. Her two arguments therefore run in opposite directions — the organism argues for fluconazole, the hardware argues for an echinocandin — and no one has yet ordered the echinocandin MIC that would settle which of them is load-bearing in her.
In the ICU, weighing a lifeline against a source of infection
This catheter is her only functioning dialysis access, with nothing readily available to replace it for at least one to two weeks. I want to treat through the line with an agent that actually works against biofilm and reserve removal for confirmed treatment failure, not pull it today and leave her without a way to dialyze. I know parapsilosis cuts against micafungin on paper — but the species argument is about the organism in her blood, and the biofilm argument is about the organism on her catheter, and we are proposing to leave the catheter in.
The evidence behind "remove when feasible" specifically found worse outcomes, including mortality, when infected lines are retained — that's not a general caution, it's a described finding. Leaving this one in place risks endocarditis or metastatic seeding in a patient this vulnerable. I'd push for removal now with a temporary non-tunneled bridge, accepting the added handling risk as the smaller one.
"No readily available replacement" describes the permanent-access timeline — it doesn't mean no bridge option exists at all.
Most of the catheter-removal-benefit data comes from populations where alternative access is genuinely trivial — a peripheral IV, a different central line site — which doesn't describe her situation. Rather than guessing which of you is right today, I'd propose a genuinely time-limited answer: 72 hours with repeat cultures at 48 to 72 hours, catheter removed immediately with a temporary bridge if cultures remain positive past that point or she shows any sign of worsening.
But I won't let the species question ride on the biofilm argument alone. IDSA names parapsilosis as one of exactly two species where echinocandin susceptibility should be tested rather than assumed, and recommends fluconazole for confirmed parapsilosis infection — so continuing micafungin is defensible only under the guideline's own narrow condition, which is a patient who is clinically improving with clearing cultures. She is afebrile and stable, so she currently meets it. Send the echinocandin MIC today. If it comes back elevated, the 72-hour window is not a trial of the catheter, it's a trial of the wrong drug — and then the answer is fluconazole plus removal, not more time.
Agreed: a 72-hour trial of micafungin with antifungal catheter lock therapy, an echinocandin MIC sent on the current isolate the same day, repeat blood cultures at 48-72 hours, and immediate catheter removal with a temporary non-tunneled bridge if cultures remain positive at that point, if the MIC returns elevated, or if she shows any clinical worsening before then.
Resolved for this admission — all three specialties signed off on the compromise timeline. The nephrologist noted for the record that she would want the trial window extended further if the 72-hour cultures clear, rather than treating clearance as an automatic endpoint to source-control vigilance.