Prospective comparison of (1,3)-beta-D-glucan detection using colorimetric and turbidimetric assays for diagnosing invasive fungal disease.
Serum (1→3)-β-D-glucan (BDG), an pan fungal antigen, is detected in some invasive fungal diseases (IFDs). We compared two commercial kits, the Fungitell assay (FA) (colorimetric) and the Wako assay (WA) (turbidimetric) over a 4-month period to prospectively test 171 patients who mainly had hematological conditions (62%) and experienced episodes (n = 175) of suspected invasive fungal infection. Twenty-three episodes due to BDG-producing fungi were diagnosed (pneumocystosis, n = 12; invasive aspergillosis, n = 5; candidemia, n = 3; invasive fusariosis, n = 2; hepato-splenic candidiasis, n = 1). Both assays provided similar areas under the curves (AUC = 0.9). Using the optimized positivity thresholds (≥120 pg/ml for FA and ≥ 4 pg/ml for WA), the sensitivity and specificity were 81.8% (CI95: 61.5-92.7), 94.8% (90.1-97.3) for FA and 81.8% (61.5-92.7), 95.4% (90.9-97.8) for WA. Negative predictive value was 97.3% (93.3-99.0) for both tests. If the manufacturer's positivity threshold (≥11 pg/ml) was applied, the WA sensitivity decreased to 50%. Among 71 patients with bacterial infections, 21.1% were FA-positive and 5.6% were WA-positive (p < 10-2). The WA performed similarly as compared to the FA with an optimized cutoff value. The WA is a single sample test that is clinically relevant when a prompt therapeutic decision is required.