Diagnostic Performance of (1→3)-β-D-Glucan Alone and in Combination with Aspergillus PCR and Galactomannan in Serum of Pediatric Patients after Allogeneic Hematopoietic Stem Cell Transplantation

Abstract

Data on biomarker-assisted diagnosis of invasive aspergillosis (IA) in pediatric patients is scarce. Therefore, we conducted a cohort study over two years including 404 serum specimens of 26 pediatric patients after allogeneic hematopoietic stem cell transplantation (alloSCT). Sera were tested prospectively twice weekly for Aspergillus-specific DNA, galactomannan (GM), and retrospectively for (1→3)-β-D-glucan (BDG). Three probable IA and two possible invasive fungal disease (IFD) cases were identified using the European Organization for Research and Treatment of Cancer and the Mycoses Study Group (EORTC/MSGERC) 2019 consensus definitions. Sensitivity and specificity for diagnosis of probable IA and possible IFD was 80% (95% confidential interval (CI): 28-99%) and 55% (95% CI: 32-77%) for BDG, 40% (95% CI: 5-85%) and 100% (95% CI: 83-100%) for GM, and 60% (95% CI: 15-95%) and 95% (95% CI: 75-100%) for Aspergillus-specific real-time PCR. However, sensitivities have to be interpreted with great caution due to the limited number of IA cases. Interestingly, the low specificity of BDG was largely caused by false-positive BDG results that clustered around the date of alloSCT. The following strategies were able to increase BDG specificity: two consecutive positive BDG tests for diagnosis (specificity 80% (95% CI: 56-94%)); using an optimized cutoff value of 306 pg/mL (specificity 90% (95% CI: 68-99%)) and testing BDG only after the acute posttransplant phase. In summary, BDG can help to diagnose IA in pediatric alloSCT recipients. However, due to the poor specificity either an increased cutoff value should be utilized or BDG results should be confirmed by an alternative Aspergillus assay.

Keywords: Aspergillus; beta-D-glucan; galactomannan; pediatric; real-time PCR.

Figure 1

Receiver operating characteristic (ROC) curve analysis of BDG testing showing the optimal cutoff value of 306 pg/mL including probable IA cases and not classified patients as controls. Sensitivity was 100%, specificity was 90%, and the fraction of the correctly classified patients was 91%. The area under the ROC curve was 0.9667 (95% CI: 0.8874, 1).

Figure 2

Temporal correlation of positive biomarkers (PCR, GM and BDG) in probable IA patients. The scale represents days in correlation to positive high resolution computed tomography (HRCT) scan; positive tests are indicated by crosses on red background. Green background indicates negative tests. Light red background (K210, BDG) indicates a serum with an indeterminate BDG result: 60–79 pg/mL; negative numbers indicate time before HRCT.

Figure 3

Medical history of a patient with several suspected pulmonary infections. Graph displays the course of C-reactive protein (CrP, red) and neutrophils counts (blue) over time, together with the results of the GM, PCR, and BDG assays (bars above the graph). For GM, PCR, and BDG assays, “red” indicates a positive test result, “green” a negative test result, “light red” an indeterminate result (only for BDG). Antifungal prophylaxis/treatment is outlined in yellow bars below the graph.