Utility of Serial β-d-Glucan Levels in Patients with High Risk for Invasive Candidiasis: A Potential Tool for Antifungal Stewardship

BACKGROUND: Invasive candidiasis (IC) is a severe infection in which diagnosis is challenging and often made late in the course of infection. Patients with delayed initiation of antifungals have high mortality risk; physicians tend to start empiric therapy at earliest clinical suspicion of IC. Excessive use of antifungals worsens selection pressure for resistance. Thus, alternative ways to aid antifungal stewardship are highly relevant. We aimed to evaluate performance of (1–3)-β-d-glucan (BDG) serial testing for antifungal stewardship to improve antifungal prescribing and to stop unnecessary use without compromising care. METHODS: This was a prospective observational study on patients at high risk of IC. Adults with recent intra-abdominal surgery, admitted to surgical intensive care unit (ICU), and prescribed an antifungal for suspected IC were included. Blood samples were taken at start of and days 3, 7, 10, 14, and weekly thereafter until antifungal is stopped, for BDG quantification with Fungitell assay. Medical records were reviewed for patient characteristics, antifungal regimen and outcomes. BDG was evaluated against clinical and microbiological outcomes. Sensitivity, specificity, positive and negative predictive values of BDG and Candida score were evaluated. RESULTS: We included 15 patients and 74 BDG levels. Patients with confirmed IC from cultures had a median BGD of >500 pg/mL and candida score of 3, compared with 55.5 pg/mL and score of 2 in those without confirmed IC. BGD assay anticipated diagnosis of IC with a sensitivity and specificity of 100% and 66.7%, with a positive and negative predictive value of 62.5% and 100% respectively. Of the five patients with confirmed IC, two had declining BDG, corresponding to clinical response to therapy. Their BDG were <80 pg/mL on day 7 and 14 of therapy, respectively, and were disharged from ICU, but one later had septic shock with Klebsiella pneumoniae bacteremia and demised. Repeat fungal cultures were negative. The remaining three had persistently high BGD of >500 pg/mL and eventually demised. No obvious trend was observed in those without confirmed IC. CONCLUSION: We were able to characterise BDG levels in patients at high risk of IC. There is utility in BGD serial testing as a tool for antifungal stewardship, however more data is required to confirm findings. DISCLOSURES: All authors: No reported disclosures.