273. Low Positive Predictive Value of β-d-Glucan in Hematology Patients Receiving Antimold Prophylaxis

Abstract Background Detection of β-D-glucan (BDG) in serum is recognized as the mycological evidence in the diagnosis of invasive fungal infection (IFI). However, its diagnostic value in low prevalence of IFI has not been elucidated. We aimed to examine the performance of BDG in hematology patients receiving antimold prophylaxis. Methods We retrospectively reviewed all BDG results performed for the purpose of diagnosis or surveillance for IFI in hematology patients receiving posaconazole or micafungin prophylaxis from January 2017 to February 2019 in a tertiary hospital. At least two consecutive positive results of BDG were regarded as positive BDG. All the episodes were classified into true-positive (TP, positive BDG with probable/proven IFI), true-negative (TN, negative BDG without probable/proven IFI), false-positive (FP, positive BDG without probable/proven IFI), false-negative (FN, negative BDG with probable/proven IFI), and nonevaluable (could not be determined for the occurrence of breakthrough IFI). When BDG test was performed in the setting of persistent fever ≥72 hours in spite of broad-spectrum antibiotics or with a suspicion of IFI, it was defined as a diagnostic BDG episode, while others were defined as a surveillance BDG episode. Results Of a total of 140 episodes, 24 episodes were non-evaluable. Among 116 evaluable episodes, 75 received induction chemotherapy for acute leukemia or myelodysplastic syndrome, 35 underwent stem cell transplantation, and 10 had intensive treatment for graft-vs.-host disease. There were three episodes of probable/proven IFI (2.6%). Ninety-one (78.4%) were performed with diagnostic purpose, while 25 (21.6%) were performed for surveillance. TP, TN, FP, and FN were 2 (1.7%), 91, 22, and 1, respectively. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value were 66.7%, 80.5%, 8.3% and 98.9%, respectively. PPV was 13.3% and 0% in diagnostic and surveillance BDG episodes, respectively. Conclusion The PPV of BDG was low in hematology patients receiving antimold prophylaxis, even in the diagnostic-driven episodes. The routine screening of BDG is not helpful, and the BDG test may be used for exclusion of IFI rather than for diagnosis in these patients. Disclosures All authors: No reported disclosures.