Invasive Fungal Infections in Critical Care: Diagnosis and Antifungal Management

Overview and Epidemiological Context

Invasive fungal infections (IFI) carry significant morbidity and mortality in the ICU setting. Invasive candidiasis accounts for the majority of ICU-acquired IFI, with crude mortality from candidaemia approaching 40-50% in critically ill patients. Invasive pulmonary aspergillosis (IPA) historically affected haematological patients but is increasingly recognised in non-neutropenic ICU patients - particularly those receiving corticosteroids, those with severe influenza or COVID-19, chronic liver disease, and those with structural lung disease. Understanding the diagnostic tools and therapeutic options for both organisms is essential CICM Final knowledge.

Diagnosis of Invasive Candidiasis

Clinical Risk Stratification: The Candida Score

The Candida score was developed to identify non-neutropenic ICU patients at high risk of invasive candidiasis who warrant empirical antifungal treatment. It integrates clinical variables that independently predict invasive infection (as opposed to colonisation).

A score $\geq 3$ is considered high risk and typically used to guide empirical antifungal initiation. In isolation, colonisation indices (ratio of colonised sites to cultured sites $> 0.5$) suggest heavy fungal burden but have lower specificity for invasive disease.

Additional clinical risk factors warranting vigilance include: - Prolonged broad-spectrum antibiotic use (> 5-7 days) - Central venous catheters, especially femoral lines - Renal replacement therapy - Prolonged ICU stay (> 7 days) - Immunosuppression including high-dose steroids - Recent abdominal surgery, anastomotic leaks, or recurrent GI perforation - Solid organ transplantation

Blood Cultures

Blood cultures remain the gold standard for diagnosis of candidaemia, but sensitivity is only approximately 50% for invasive candidiasis (deep-seated infection without fungaemia may be culture-negative). Time to positivity is typically 24-72 hours. Optimal technique requires: - Multiple sets from separate sites - Adequate volume (10 mL per bottle) - Dedicated fungal lysis-centrifugation bottles where available

Once positive, species identification guides antifungal selection - Candida auris, C. krusei (now Pichia kudriavzevii), and emerging C. glabrata (now Nakaseomyces glabratae) have variable azole susceptibility and specific management implications.

Beta-D-Glucan (BDG)

$\beta$-1,3-D-glucan is a cell wall polysaccharide released by most pathogenic fungi (including Candida and Aspergillus, but notably not Cryptococcus or Mucorales, which have capsules or lack significant glucan exposure).

BDG is best interpreted in clinical context, not in isolation. Serial negative values (two consecutive) carry a high negative predictive value (>90%) and can support de-escalation decisions.

Diagnosis of Invasive Aspergillosis

Galactomannan (GM)

Galactomannan is a polysaccharide component of the Aspergillus cell wall released during hyphal growth. Serum GM is the primary biomarker for IPA.

BAL GM has superior sensitivity to serum GM in non-neutropenic ICU patients and should be obtained at bronchoscopy whenever IPA is suspected.

Combined BDG and GM testing increases sensitivity - a positive result for both in the right clinical context is highly supportive of IPA.

HRCT Features of IPA

CT thorax is indispensable in suspected IPA. Key patterns include:

The halo sign is time-sensitive - it may only persist for a few days in its classic form before the surrounding oedema resolves or consolidation develops. Early CT acquisition (within 24-48 hours of clinical suspicion) is critical.

Diagnostic Classifications

The EORTC/MSG criteria define IFI as proven, probable, or possible based on host factors, mycological criteria (culture, GM, GM-BAL), and clinical/radiological criteria. In the ICU context, most diagnoses will be "probable" - the treating intensivist must act on this without awaiting histological proof.

Antifungal Pharmacology: Mechanisms and Spectrum

Echinocandins

Echinocandins are large cyclic lipopeptides that inhibit $\beta$-1,3-glucan synthase - specifically the Fks1p catalytic subunit - thereby disrupting fungal cell wall biosynthesis. Without structural glucan, the cell wall loses integrity and the cell lyses.

Dosing summary:

Echinocandins have relatively few significant drug interactions and are generally well tolerated, making them preferred agents in the polypharmacy-heavy ICU patient.

Azoles

Voriconazole is the agent of choice for invasive aspergillosis and many non-Mucorales mould infections. It inhibits fungal cytochrome P450 enzyme lanosterol 14-$\alpha$-demethylase, preventing ergosterol synthesis.

Therapeutic drug monitoring (TDM) is mandatory for voriconazole in ICU patients due to highly variable pharmacokinetics.

Isavuconazole offers an alternative with activity against Aspergillus and Mucorales, better tolerability than voriconazole, and - uniquely - causes QTc shortening rather than prolongation, making it contraindicated in familial short QT syndrome. IV and oral formulations are available. Loading: 372 mg (as isavuconazonium sulfate) every 8 hours for 6 doses, then 372 mg once daily.

Posaconazole has broad activity including Mucorales and is particularly valuable as prophylaxis in haematological patients. Serum trough target is 0.5-1.5 mcg/mL for prophylaxis and > 1.0 mcg/mL for treatment of mould infections.

Amphotericin B

Binds ergosterol in the fungal cell membrane causing pore formation and cell death. Broadest spectrum of any antifungal - active against most Candida, Aspergillus, Mucorales, and Cryptococcus.

Liposomal amphotericin B remains a cornerstone agent for mucormycosis (alongside surgical debridement) and for infections not amenable to azole or echinocandin therapy.

Empirical, Pre-emptive, and Targeted Antifungal Strategies

Empirical antifungal therapy for suspected invasive candidiasis: - First-line: echinocandin (caspofungin, micafungin, or anidulafungin) - Rationale: fungicidal, excellent Candida coverage including non-albicans species, low toxicity, few interactions - Fluconazole acceptable only if low severity, no recent azole exposure, and low risk of C. glabrata or C. krusei

Targeted therapy for invasive aspergillosis: - First-line: voriconazole IV/oral with TDM - Alternative: isavuconazole (particularly if QTc concerns, renal impairment precluding IV voriconazole vehicle cyclodextrin, or Mucorales co-infection concern) - Salvage: liposomal amphotericin B, or combination therapy in refractory cases - Remove immunosuppression where possible; consider surgical debridement for focal lesions or haemoptysis

De-escalation Strategy

De-escalation is a core stewardship principle - antifungals carry significant toxicity, promote resistance (particularly echinocandin resistance in C. glabrata), and are costly.

De-escalation in Candidaemia

Duration of treatment: minimum 14 days from first negative blood culture and resolution of symptoms, with evidence of no deep-seated infection (ophthalmology review mandatory - endophthalmitis occurs in ~5-10%).

De-escalation in Invasive Aspergillosis

• IV to oral voriconazole switch when patient can absorb enteral medications and is clinically improving
• Oral bioavailability of voriconazole is ~96% - IV-to-oral switch is pharmacologically sound
• Minimum treatment duration: 6-12 weeks - determined by degree of immunosuppression, extent of disease, and radiological response
• Serial GM monitoring can guide response - falling ODI values correlate with treatment success
• Repeat CT at 4-6 weeks: stable or improving lesions support continuation; new lesions or cavitation requiring surgical evaluation

Antifungal Stewardship Principles

• Two consecutive negative BDG values in a stable patient on empirical therapy supports stopping antifungals (NPV > 90%)
• Avoid empirical antifungal therapy in patients with a Candida score < 3 without microbiological support
• De-escalate from broad-spectrum (e.g., amphotericin) to targeted agent as soon as species and susceptibility known
• Review duration at each ward round - no "auto-continue" for antifungals

CICM Final Implications

Hot Case / Viva Approach

In a viva on an ICU patient with unexplained fever and septic shock unresponsive to antibiotics, a structured approach is expected:

1. Identify risk factors systematically: TPN, abdominal surgery, prolonged CVC, broad-spectrum antibiotics, immunosuppression, RRT
2. Enumerate diagnostic workup: blood cultures (minimum 2 sets, fungal media), BDG, GM-BAL if respiratory involvement/infiltrates on imaging, CT chest (halo sign?), urine MC&S, ophthalmology consult if candidaemia suspected
3. Empirical decision: Candida score $\geq 3$ → start echinocandin now; do not wait for culture results; justify choice over fluconazole
4. Species-driven adjustment: articulate when to step down (susceptible C. albicans → fluconazole) versus maintain or escalate (C. auris → echinocandin; refractory IPA → L-AmB)
5. Organ support integration: echinocandins require no renal dose adjustment - safe in AKI/CRRT; voriconazole IV contains cyclodextrin vehicle which accumulates in renal failure → prefer oral voriconazole or switch to isavuconazole when GFR < 50 mL/min
6. Stewardship: articulate triggers for stopping empirical therapy (clinical improvement, two negative BDG, cultures negative at 5 days)
7. Resistant organisms: C. auris clusters require infection control escalation - contact precautions, environmental decontamination; notify infection control immediately

Key Numbers to Recall

The overarching principle for the CICM Final is to demonstrate understanding of why - why echinocandins first (fungicidal, safe, broad Candida cover), why voriconazole for Aspergillus (superior evidence, excellent bioavailability), and why de-escalation matters (toxicity, resistance, cost, stewardship).