Medical Mycology

Educational study digest pitched at a new ID fellow. Clinical recommendations trace to the cited guidelines and trials; the enrichment box is reviewer synthesis of additional evidence, each claim cited. Not medical advice, and not a substitute for the primary guidelines.

◆Candidemia: why the answer starts with an echinocandin — and when it stops

For candidemia and invasive candidiasis the reflex first move is an echinocandin (micafungin, caspofungin, anidulafungin), and the reasons are as much practical as microbiological. Candida species are reliably susceptible; the drugs carry essentially no meaningful drug interactions and need no renal dose adjustment, which makes them easy to run in a sick, poly-pharmacy ICU patient. But the first thing to say out loud is source control, source control, source control — drain the collection, pull the line, wash out the abdomen — because no antifungal salvages an undrained focus.

The convenience comes from the same property that limits it: echinocandins are highly protein-bound large molecules, and that binding keeps them out of two compartments that matter. They do not achieve useful concentrations in the CNS or the eye. So Candida meningitis or endophthalmitis is not an echinocandin problem — those go to an agent that penetrates: fluconazole or voriconazole in the CNS, and amphotericin B where potency is needed. Critical illness bends the pharmacokinetics further: an expanded volume of distribution and altered albumin binding in the critically ill can pull echinocandin exposure down at the site of infection, which is part of why some reach for amphotericin B in the most unstable patients.

The species that forces a second thought is Candida parapsilosis, which characteristically carries higher echinocandin MICs than the other Candida. In documented C. parapsilosis disease an echinocandin may under-perform, and a susceptible-tested azole is often the better-matched drug. Two related habits keep a fellow out of trouble: distinguish colonization from infection before treating at all — a Candida-positive nonsterile wound or drain, in a patient without host risk, sterile-site growth, tissue invasion, or corroborating imaging/biomarkers, may not need therapy — and remember that echinocandin resistance can emerge on therapy, so a clinical relapse warrants repeat susceptibilities.

Once the fire is out, step down from IV echinocandin to an oral azole. The 2025 global candidiasis guidance sets the bar: switch after at least 5 days of an echinocandin only when the patient is hemodynamically stable, blood cultures have cleared, the patient is non-neutropenic, source control has been achieved, the patient can tolerate oral intake, and susceptibility is confirmed. An earlier transition is reasonable for species with a low likelihood of azole resistance — C. albicans, C. parapsilosis, and C. tropicalis.

◆Antifungal resistance: two enzymes and a short list of names

Two mechanisms cover most of what gets asked. Azole resistance most classically runs through the target enzyme: a mutation in CYP51A, the gene encoding C14-α-demethylase — the lanosterol-demethylating step the azoles inhibit, and the mechanism to know for both Candida and Aspergillus. The second azole route is drug efflux, pumps that throw the azole back out of the cell, the flavor most associated with C. glabrata. Echinocandin resistance has its own address: mutations in FKS1 and FKS2, the genes for β-(1,3)-glucan synthase — the very target the echinocandins block.

The other half of the question is a memorized list. When the stem says “fluconazole-resistant Candida,” think C. krusei, C. auris, C. haemulonii, and some C. glabrata. C. krusei is intrinsically fluconazole-resistant; C. auris is the emerging multidrug-resistant species that also drives infection-control worry; and C. glabrata is the variable one, with dose-dependent susceptibility and efflux-mediated resistance — which is why it sits on the list with an asterisk rather than as an absolute.

◆Azole toxicities the vignette is built around

A handful of azole toxicities are tested precisely because the drug, not a new disease, is the answer. Voriconazole is the azole that causes photosensitivity, and with chronic use that photo-injury extends to fluoride-associated periostitis — diffuse bone pain with periosteal reaction in a patient on long-term voriconazole. It is also the azole with the metabolism trap: clearance runs largely through CYP2C19, and patients of Asian descent are disproportionately slow metabolizers, so standard doses can run supratherapeutic and drive the neurologic and visual effects voriconazole is known for. As a class, the azoles are the antifungals to fear for drug interactions — potent CYP effects that swing the levels of co-administered drugs.

Isavuconazole is the azole to file under the QT interval, and the discriminator is direction: unlike the QT prolongation typical of voriconazole, fluconazole, and posaconazole, isavuconazole runs the other way and is associated with QT shortening. A broad-spectrum azole that shortens rather than lengthens the QT is exactly the kind of reversal a question is built to reward.

◆Cryptococcosis: reading the antigen and staging the therapy

Cryptococcal antigen (CrAg) is the workhorse, and the two assays are not interchangeable. The lateral flow assay (LFA) detects the polysaccharide capsule with gold-conjugated monoclonal antibodies, brings the detection threshold down to roughly 1–2 ng/mL, returns in about 10 minutes, and — importantly — is less susceptible to the prozone effect, the paradoxical false-negative that very high antigen loads can produce. The older latex agglutination assay uses antibody-coated latex beads, detects only down to about 50 ng/mL, and takes roughly 2 hours. The number to carry: cryptococcemia or a high fungal burden — a CrAg titer ≥ 1:512 — should be managed like CNS disease even without a positive CSF, because that burden predicts dissemination.

Treatment is staged in three phases. Induction runs about 2 weeks on liposomal amphotericin B plus flucytosine (5-FC), the combination that most rapidly sterilizes the CSF; in low-resource settings a single high-dose amphotericin strategy is an accepted alternative (see the enrichment box). Consolidation follows for about 8 weeks on fluconazole 400–800 mg daily, and maintenance (secondary prophylaxis) continues on fluconazole 200 mg daily for at least 12 months, or until immune reconstitution. Amphotericin + flucytosine to induce, then fluconazole to consolidate and maintain — that sequence is the spine of every cryptococcal-meningitis answer.

◆Invasive moulds: the halo sign, and the ones amphotericin can’t touch

In a neutropenic patient — the classic being a stem-cell transplant recipient who has failed to engraft with an ANC of zero — a new pulmonary nodule with a surrounding rim of ground-glass, the halo sign, is invasive aspergillosis until proven otherwise. The physiology explains the picture: Aspergillus is angioinvasive, invading and thrombosing blood vessels, so the dense center is fungus and infarcted lung while the halo is hemorrhage bleeding into the surrounding alveoli. It is highly suggestive rather than absolutely diagnostic, and the setting sharpens it — a patient on fluconazole prophylaxis is not protected against a mould, and Cryptococcus (which fluconazole would cover, and which does not make a halo) drops down the list.

The trap arrives when the BAL shows septate, branching hyphae but the mould is one amphotericin cannot touch. Scedosporium (S. apiospermum) is intrinsically resistant to amphotericin B, and the answer is to treat with voriconazole — FDA-approved as salvage therapy for S. apiospermum, available intravenously, and the preferred agent; echinocandins are not adequate alone. Keep the whole amphotericin-resistant roster together, because questions rotate through it: Aspergillus terreus (the Aspergillus species that, unlike A. fumigatus, resists amphotericin), Scedosporium species, and Trichosporon — and, among the yeasts, Candida lusitaniae. In each, a clue pointing to one of those names should redirect you off amphotericin and toward a triazole.

◆Dimorphic fungi: geography, tissue form, culture form

When a cavitary or nodular pulmonary lesion in the right host turns out to be a thermally dimorphic fungus, three facts pin the species — where the patient was, what the yeast (tissue) form looks like at body temperature, and what the mould (culture) form looks like grown cooler. The differential for infectious cavitary pneumonia is broad — pyogenic abscess, Nocardia, Actinomyces, Rhodococcus, and melioidosis; TB and NTM (MAC, M. kansasii); Cryptococcus, invasive moulds (Aspergillus, Mucorales), and the dimorphics — but the three endemic dimorphs separate cleanly on the grid below.

The tissue form is usually the fastest discriminator on a slide: intracellular clusters point to Histoplasma, giant endospore-filled spherules to Coccidioides, and a broad-based budding yeast to Blastomyces. Histoplasmosis is the one that most often cavitates in a compromised host — the tried-and-true approach for progressive disease is amphotericin B followed by an oral azole (itraconazole, with posaconazole as an alternative).

◆Fungal prosthetic joint infection: remove the hardware, load the cement

Fungal prosthetic joint infection is uncommon enough that the evidence is thin — the optimal antifungal and duration are genuinely under-defined — but two principles carry. Source control means hardware removal, because a biofilm-laden implant will not clear on drugs alone; and local delivery via antifungal-laden cement, classically voriconazole in the spacer cement, supplements systemic therapy. The full construct — staged exchange, systemic durations, and the bacterial-PJI logic it borrows from — lives in the companion Bone & Joint Infections digest.

Sources: Global guideline for the diagnosis and management of candidiasis (ECMM/ISHAM/ASM; Lancet Infect Dis 2025; doi:10.1016/S1473-3099(24)00749-7); Global guideline for the diagnosis and management of cryptococcosis (ECMM/ISHAM; Lancet Infect Dis 2024; doi:10.1016/S1473-3099(23)00731-4); Cryptococcal Disease in Diverse Hosts (N Engl J Med 2024; doi:10.1056/NEJMra2311057; PMID 38692293); AMBITION single-dose liposomal amphotericin B for cryptococcal meningitis (N Engl J Med 2022; doi:10.1056/NEJMoa2111904; PMID 35320642); spectrum of antifungals for common molds and dimorphic fungi (Johnson et al., Infect Dis Clin North Am 2025).