Antifungal resistance in Candida glabrata and Candida krusei is primarily intrinsic or rapidly acquired resistance to fluconazole, driven by efflux pump overexpression and target-site mutations, and it is the main reason these two species cause breakthrough infections during azole prophylaxis. C. krusei is intrinsically fluconazole-resistant in essentially 100% of isolates. C. glabrata is dose-dependent susceptible in roughly 50 to 70% of isolates but acquires fluconazole resistance in 15 to 25% of cases during prolonged exposure. Echinocandins remain first-line therapy for both species, with resistance emerging in 2 to 5% of C. glabrata isolates in high-exposure settings.
Why These Two Species Resist Azoles Differently
The mechanism is the decisive difference between the two organisms. Understanding it explains why one resists from the first exposure and the other gradually becomes resistant under pressure.
Candida krusei is intrinsically resistant to fluconazole. The mechanism rests partly on a low-affinity target: its ERG11-encoded lanosterol 14-alpha-demethylase binds fluconazole poorly. Baseline expression of the ABC transporter CDR1 is also elevated compared with C. albicans. Because resistance is present before drug exposure, no selection step is required. The 2022 CLSI M27M44S breakpoint definitions classify C. krusei as inherently non-susceptible to fluconazole, so susceptibility testing is not even required to justify avoiding it.
Candida glabrata works differently. It is haploid, unlike the diploid C. albicans, which means a single point mutation in a resistance gene produces an immediate phenotype rather than requiring two alleles. This haploidy is a structural reason C. glabrata acquires resistance faster than most Candida species. The principal mechanism is upregulation of the CDR1 and CDR2 genes, which encode ATP-binding cassette efflux pumps that physically expel azole molecules from the cell. A secondary mechanism is point mutation in ERG11, most commonly at amino acid positions Y132F, K143R, and S405F, which reduce target binding.
In plain terms: C. krusei is born resistant, while C. glabrata learns resistance under drug pressure, and its single-copy genome makes that learning fast.
Quantified Resistance Rates and Species Distribution
The scale of the problem has shifted measurably in the past decade. Candida glabrata is now the second most common cause of candidemia in the United States after C. albicans, accounting for roughly 15 to 25% of bloodstream isolates in surveillance data. C. krusei is less common, at approximately 2 to 4% of candidemia isolates, but carries the highest crude mortality of the common non-albicans species.
Fluconazole resistance rates track exposure history. In azole-naive patients, C. glabrata fluconazole resistance sits near 10 to 15%. In patients with prior azole exposure or in hematology-oncology units, that figure rises to 30 to 45%. The CDC's 2019 Antibiotic Resistance Threats report classified fluconazole-resistant Candida as a serious threat, estimating 34,000 cases and 1,700 deaths annually in the US across all Candida species. Echinocandin resistance in C. glabrata remains lower but is rising, reported at 2 to 5% overall and as high as 10 to 15% in centers with heavy prior echinocandin use.
The numbers below reconcile into a single practical statement: intrinsic resistance in C. krusei approaches 100%, acquired fluconazole resistance in C. glabrata ranges from roughly 10% to 45% depending on prior exposure, and echinocandin resistance in C. glabrata sits below 15% everywhere reported but is trending upward.
Comparison: C. glabrata Versus C. krusei
| Feature | Candida glabrata | Candida krusei |
|---|---|---|
| Fluconazole susceptibility | Dose-dependent in 50-70%, resistant in 10-45% | Intrinsically resistant, ~100% |
| Primary resistance mechanism | Efflux pump upregulation (CDR1/CDR2), ERG11 mutation | Low-affinity ERG11 target, elevated CDR1 |
| Genome ploidy | Haploid | Diploid |
| First-line therapy | Echinocandin | Echinocandin |
| Echinocandin resistance (reported) | 2-5% (up to 15% in high-use centers) | Under 2%, rare |
| Amphotericin B role | Salvage therapy | Active in vitro; used when echinocandin unavailable |
| Typical setting | Elderly, ICU, prior azole use | Hematologic malignancy, neutropenia |
| Crude mortality (candidemia) | 30-40% | 40-50% |
Timeline of Resistance Development and Clinical Course
Resistance is not instantaneous in C. glabrata. It develops along a traceable trajectory that maps to exposure duration. The timeline below reflects the typical clinical sequence in a patient receiving antifungal therapy.
| Timepoint | Biological Event | Clinical Implication |
|---|---|---|
| Day 0 | Colonization with susceptible C. glabrata | No symptoms, no treatment indicated |
| Days 1-7 | Azole prophylaxis or therapy begins | Selection pressure on efflux variants starts |
| Weeks 2-4 | CDR1 upregulation emerges in subpopulation | MIC rises from 8-16 mg/L toward 32-64 mg/L |
| Weeks 4-8 | Stable fluconazole resistance established | Breakthrough candidemia possible on azole therapy |
| Weeks 6-12 | Echinocandin switch initiated | FKS1 hotspot mutations may appear under pressure |
| Months 3-6 | Echinocandin resistance in 2-5% of cases | Requires amphotericin B salvage |
| Months 6+ | Multidrug-resistant phenotype possible | Limited options, higher mortality |
The key inflection is the transition from colonization to breakthrough infection. A patient who develops C. glabrata candidemia while already on fluconazole has a substantially higher probability of a fluconazole-resistant isolate than one who is azole-naive. Per the 2016 IDSA clinical practice guideline for candidiasis, this exposure history should directly drive empiric therapy choice before susceptibility results return, which can take 24 to 72 hours.
Misconceptions and Non-Obvious Mechanism Details
Three beliefs about these species are common and incorrect.
First, many clinicians assume that a fluconazole MIC in the susceptible-dose-dependent range means the drug will work if the dose is increased. In C. glabrata, dose escalation to 800 mg daily can overcome moderate efflux, but not target-site ERG11 mutations, and it fails entirely once the MIC exceeds 64 mg/L. The dose-dependent category is not a guarantee; it is a window that closes as resistance accumulates. The 2022 CLSI breakpoints now use a single susceptible-dose-dependent category for fluconazole, which has reduced interpretive ambiguity but did not change the underlying biology.
Second, the source of an infection is often assumed to be exogenous. Molecular typing shows that the majority of C. glabrata candidemia arises from the patient's own colonizing gastrointestinal or genitourinary flora after antibiotic or azole pressure selects a resistant subclone. This changes infection control logic: the reservoir is usually the patient, not a shared environmental source, though clonal outbreaks do occur.
Third, echinocandins are often treated as uniformly reliable against these species. They target the FKS1 subunit of beta-1,3-D-glucan synthase, and resistance requires only a single hotspot mutation in FKS1, at positions S629, F659, or S663. Because C. glabrata is haploid, one mutation confers full resistance, whereas in diploid species two mutations or a dominant allele are required. This is a counterintuitive point: the same ploidy that accelerates azole resistance also accelerates echinocandin resistance.
What to Actually Do: Practical Clinical Actions
Concrete actions follow from the mechanisms above.
- Obtain species-level identification and susceptibility testing for every non-albicans Candida bloodstream isolate; do not rely on empiric azole therapy.
- Treat known C. krusei with an echinocandin as first-line; avoid fluconazole without exception.
- For C. glabrata candidemia, start an echinocandin and de-escalate only after susceptibility confirms fluconazole susceptibility, per the 2016 IDSA guideline.
- Review prior azole exposure in every patient with candidemia; a positive history shifts the pretest probability of resistance toward 30 to 45%.
- Reserve amphotericin B for echinocandin-resistant or multidrug-resistant isolates and for cases where echinocandin is contraindicated.
- Screen high-risk hematology patients for colonization; colonization with C. glabrata or C. krusei raises the risk of subsequent invasive disease substantially.
Red-flag warning signs that demand immediate action include: candidemia persisting more than 72 hours on appropriate therapy, breakthrough infection while on an echinocandin, prior fluconazole exposure within 30 days, and any isolate with a fluconazole MIC above 64 mg/L.
Limits of Current Evidence and Unresolved Questions
Several claims in this field rest on surveillance data rather than controlled trials. The precise attributable mortality of resistance itself, independent of host factors, is difficult to isolate because resistant infections cluster in sicker patients. Reported resistance rates vary by geography: northern European surveillance typically reports lower C. glabrata azole resistance than US and southern European series, likely reflecting different azole prescribing volumes. Rapid molecular panels now detect ERG11 and FKS1 mutations in 1 to 4 hours, but their clinical impact on mortality has not been established in randomized trials.
One hypothesis, not yet confirmed, is that gut microbiome disruption from broad-spectrum antibiotics directly promotes expansion of resistant C. glabrata clones. The mechanism is plausible given the colonization-to-infection pathway, but causality remains unproven. Clinicians should treat resistance rates as setting-specific and local antibiograms as more reliable than national averages for individual treatment decisions.