Serum β-D-glucan (BDG) as a biomarker for, or exclusion of, invasive fungal disease in haematology and intensive care unit patients – a pilot study

Catriona Halliday*1, Menuk Jayawardena1, Katherine Garnham1, Tasnim Hassan2, Neela Joshi Rai1, Jen Kok1,2, 5, Vineet Nayyar3, David Gottlieb4, Nicky Gilroy2, 5, and Sharon Chen1,2, 5

Introduction
The detection of fungal biomarkers in blood, such as B-D-glucan (BDG), has good potential for early diagnosis, or for exclusion of I invasive fungal diseases (IFDs), but there are no data of its clinical utility in Australia 
Objectives
1)	To perform a pilot study to derive experience of serum BDG testing in an Australian setting for the first time
2)	Determine the clinical utility of serum BDG testing in the early detection (or exclusion) of IFD in high risk haematology and intensive care unit (ICU) patients

Methods
Serum from patients with proven IFD was employed to establish “positive” controls. Serum from high risk haemopoeitic stem cell transplant [HSCT] and acute leukaemia patients were collected prospectively during their neutropenic period (twice weekly as per for Aspergillus galactomannan) were tested for by BDG using the Fungitell® assay (Associates of Cape Cod Inc.). For ICU patients, serum was collected daily for the first 3 days in ICU in those with medium-to-high IFD risk including complex gastrointestinal (GI) surgery procedures.  BDG results were not used in clinical management. Patients were assessed for IFD with EORTC/MSG criteria.  
Results
A total of 126 specimens from 31 patients were studied; 18 were haematology patients and 
12 were classed as ICU patients. There was a single renal transplant patient with endovascular IFD infection. This patient had 6 samples tested over 3 months all returning a positive BDG result; Aspergillus hyphae were seen in arterial tissue. 

Of haematology patients with malignancy, 15 had received an allogeneic or autologous transplant. All 6 patients who had a consistently negative BDG test (average 3 specimens) had no evidence of IFD. Two patients had proven IFD (aspergillosis, fusariosis) and had consistently elevated BDG levels in multiple specimens (mean OD, 1994 pg/ml; range, 1742-2436 pg/ml). Three patients had probable IFD with mean ODs of 676 pg/ml and range of 163-2557 pg/ml. The 4 patients with possible IFD had intermittent positive BDG tests (mean OD, 255 pg/ml; range, 96-536 pg/ml). In general, patients with definite/probable infection had 4 times higher OD reading compared with patients with possible IFD. There were 3 patients classed as no IFD with intermittently positive BDG tests. 

Of 12 ICU patients, 38 serum specimens were studied. Positive BDG results were obtained for 7 patients (15 samples); 2 had definite IFD (candidemia and intra-abdominal candidiasis) and two had possible intra-abdominal IFD. The mean OD in patients with definite IFD was 122 pg/ml (range 103-145 pg/ml). The remaining 3 patients were considered to have no IFD (mean OD, 256 pg/ml; range 92-627 pg/ml) Five patients had no IFD and had consecutively negative results; all were receiving antifungal prophylaxis at time of sampling.

Conclusions
As expected, the serum BDG assay performed well in patients with definite/probable IFD. For patients with possible IFD, the BDG assay may have helped upgrade patients to probable IFD if the test was available. Patients with consistently negative BDG are unlikely to have IFD and in ICU, antifungal prophylaxis could have ceased. Testing of a larger cohort of our patient is essential.

Serum B-D-glucan as a biomarker for, or exclusion of, invasive fungal disease - a pilot study