The Isavuconazole Therapeutic Drug Monitoring in Critically Ill Patients

Presenter: Wu, Vicky

This retrospective single-centre study evaluated isavuconazole levels in critically ill patients, where altered pharmacokinetics may impact drug exposure despite prior evidence suggesting therapeutic drug monitoring (TDM) is not routinely required. Adults (≥20 years) admitted to the ICU with available isavuconazole levels were included, excluding samples obtained after dose adjustment.

A total of 26 patients (28 samples) were analysed. Patients were predominantly male (73%), with a median age of 58 years (IQR 47.8–68), median body weight 58 kg (51.8–63.3), and median BMI 21 (19.9–23.5). Underlying conditions included malignant hematology (61.5%), oncology (26.9%), rheumatology (23%), and organ transplant (7.6%). Median SOFA score was 13.5 (8–15.25). Hyperbilirubinemia was present in 57.1%, liver cirrhosis in 7.6%, and acute liver failure occurred in 9%.

Isavuconazole levels were measured at a median of 10 days in 23 samples (excluding 5 samples collected after >21 days of therapy). The median drug level was 2.94 mg/L, with 60.7% within the therapeutic range (2–5 mg/L). Subtherapeutic levels were observed in 17.8% of samples, including 7.1% <1 mg/L; two of these occurred during drug–drug interaction with rifabutin. Supratherapeutic levels were seen in 21.4%, with both patients with liver cirrhosis falling into this category. Additionally, 26% of levels were drawn before day 7 (median 4.5 days), with a median level of 2.05 mg/L.

These findings indicate variability in isavuconazole exposure in critically ill patients, with a substantial proportion outside the therapeutic range despite standard dosing, particularly in the presence of liver disease or drug interactions.

Survival Trends in Invasive Candidiasis in the ICU, 229 US Hospitals, 2018-2023

Presenter: Badesch, Brittany

This study evaluated whether mortality in ICU patients with invasive candidiasis (IC) has changed over time, accounting for host factors and excluding COVID-19-related confounding. Using the PINC-AI database, 5,041 adult encounters with microbiologically confirmed IC across 229 hospitals were analysed. IC was defined as Candida growth in blood or other sterile sites within −2 to +10 days of antifungal therapy initiation.

Of these, 2,532 patients (50.2%) required ICU admission. Among ICU patients, 66% received mechanical ventilation, 72% required vasopressors, and 24% were immunocompromised. Bloodstream infections accounted for 95% of cases, with 5% from other sterile sites. The predicted probability of mortality (in-hospital death or discharge to hospice) in non-COVID ICU patients was 0.36 (95% CI 0.34–0.38). Over time, there was no statistically significant change in mortality, with a difference of −0.023 (95% CI −0.0859 to 0.04) when comparing 2018 to 2023.

These findings indicate persistently high mortality in ICU patients with invasive candidiasis, with no meaningful improvement over the past six years despite advances in care.

Impact of a Pharmacy-Driven (1,3)-Β-D-Glucan Algorithm on Micafungin Duration of Therapy in the ICU

Presenter: Kheir, Daniella

This retrospective, multisite pre-post cohort study evaluated whether a (1,3)-β-D-glucan (BDG)–guided algorithm could reduce micafungin duration of therapy (DOT) compared with standard of care (SOC) in non-neutropenic critically ill adults with risk factors for invasive candidiasis. The BDG algorithm was compared with a prior-year SOC cohort.

Of 318 screened patients, 110 were included. Baseline characteristics were similar, although invasive candidiasis rates were lower in the SOC group (1.8% vs 7.3%). Overall micafungin DOT did not differ between groups [94.7 (45.0–146.3) vs 95.4 (49.2–206.4) hours; p=0.84]. However, DOT was significantly shorter when BDG results were available within 48 hours [53.0 (46.1–98.1) vs 140.7 (72.8–226.6) hours; p=0.014], and when algorithm compliance was achieved following negative BDG results [47.6 (35.6–68.8) vs 95.4 (49.2–206.4) hours; p=0.005].

The BDG assay demonstrated a negative predictive value of 96.6% and a positive predictive value of 11.5%. Dialysis patients were more likely to have false-positive results (phi=0.40, 95% CI 0.15–0.60; p=0.002). Immunocompromised and surgical patients had longer DOT. No significant differences were observed in length of stay, mortality, 30-day readmissions, or cost.

These findings suggest that while the BDG algorithm did not reduce overall antifungal use, timely test results and adherence to the algorithm were associated with shorter micafungin exposure.

Use of Topical Amphotericin for Adjunctive Management of Fungal Infections in Acute Burn Injuries

Presenter: Lee, Bernice

This case series describes the use of topical amphotericin B (AmB) as adjunctive therapy in three patients with burn-related invasive fungal infections, where systemic antifungal therapy is often limited by toxicity and pharmacokinetic challenges.

Patient A, a 30-year-old male with 70% total body surface area (TBSA) burns, developed an Aspergillus wound infection and was treated with voriconazole. Due to difficulty achieving therapeutic levels and slow wound healing, topical conventional AmB was added. Patient B, a 49-year-old male with 5% TBSA burns, developed a Rhizopus infection and received IV liposomal AmB with additional topical liposomal AmB. Patient C, a 62-year-old male with 36% TBSA burns and Aspergillus infection, was treated with voriconazole and topical conventional AmB concurrently.

Topical AmB was compounded to a concentration of 24 mcg/mL in all cases. All three patients achieved successful clinical outcomes with combination therapy, and no adverse effects related to topical treatment were reported.

These findings suggest that topical AmB may serve as a useful adjunct to systemic antifungal therapy in burn patients, potentially improving local drug delivery without added toxicity.

Incidence of Therapeutic Azole Dosing in ECMO Patients

Presenter: Feih, Joel

This retrospective cohort study evaluated the attainment of therapeutic concentrations of triazole antifungals in patients receiving extracorporeal membrane oxygenation (ECMO), where altered pharmacokinetics may impact drug exposure. Patients on ECMO receiving voriconazole or posaconazole with therapeutic drug monitoring between July 2015 and July 2025 were included.

A total of 12 patients with 36 drug level measurements were analyzed. The median age was 42 years (IQR 26–60), and 58% were male. Common indications included aspergillosis and lung transplant prophylaxis (n=4 each, 33%). Overall, 41.6% (15/36) of concentrations were within the therapeutic range, including 39% for voriconazole and 50% for posaconazole. Supratherapeutic levels occurred in 16.7% of samples following dose titrations.

Among 10 patients newly initiated on azoles during ECMO, only 4/10 achieved therapeutic levels on initial monitoring. The median daily dose required to reach therapeutic levels was 4 mg/kg (IQR 4–6.8) for voriconazole and 300 mg/day (IQR 200–500) for posaconazole.

These findings indicate that achieving therapeutic azole concentrations during ECMO is challenging, with frequent subtherapeutic levels and variable dosing requirements, particularly for voriconazole.

Empiric Antifungal Therapy for Critically Ill Patients with Candidemia

Presenter: Vashee, Chandni

This retrospective cohort study evaluated whether early empiric antifungal therapy improves mortality in patients with candidemia. The study included 399 patients, excluding those already on antifungals at the time of culture or those not treated within 5 days. Patients were compared based on antifungal initiation within 24 hours versus after 24 hours of blood culture collection.

Of the cohort, 97 patients (24.3%) received antifungals within 24 hours, while 302 (75.7%) received treatment after 24 hours; only 4.5% received therapy within 3 hours. The mean time to antifungal initiation was 39.4 hours (SD 22.0), with no differences in vital signs between groups.

Overall, 7-day mortality was 11.8% (47 patients), and 30-day mortality was 25.6% (102 patients). Among ICU patients, mortality was higher, with 7-day mortality of 23.9% versus 5.1% in non-ICU patients, and 30-day mortality of 51.4% versus 11.3%. There was no statistically significant difference in mortality between early and delayed treatment groups (30.9% vs 23.8%; p=0.16).

These findings indicate that most patients experienced delayed antifungal initiation, and early treatment within 24 hours was not associated with improved mortality, potentially reflecting differences in illness severity.

Critical Care Congress 2026, March 22-24, Chicago