Presence of Candida cell wall derived polysaccharides in the sera of intensive care unit patients: relation with candidaemia and Candida colonisation

Abstract

Introduction: Prompt diagnosis of candidaemia and invasive candidosis is crucial to the early initiation of antifungal therapy. The poor sensitivity of blood cultures (BCs) has led to the development of fungal glycan tests as a diagnostic adjunct. We analysed the performance of tests for the detection of circulating β-D-1,3-glucan (BDG) and mannan in the intensive care unit (ICU) setting.

Methods: This retrospective, case-control study included 43 ICU patients with candidaemia and 67 controls, hospitalised on the same ward and assessed weekly for yeast colonisation with simultaneous serum sampling; 340 sera taken before and after positive BCs were available for the cases group and 203 for the controls. BDG and mannan levels were determined using the Fungitell® and Platelia™ Candida Ag tests, respectively.

Results: BDG was detected early in sera from cases patients but was also present in several sera from controls. Increasing the cut-off from 80 pg/mL to 350 pg/mL and 800 pg/mL resulted in sensitivity/specificity ratios of 0.97/0.31, 0.65/0.74, 0.30/0.86, respectively. Detection of mannan was more specific but lacked sensitivity. No obvious correlation was found between BDG and colonisation, but a trend existed between high colonisation and high BDG. Candidaemia relapses were associated with a rise in BDG and mannan but, in contrast to the transient nature of mannan, BDG persisted up to 7 weeks after positive BCs.

Conclusion: A combination of mannan and BDG tests could be used to guide pre-emptive therapeutic decisions in ICU patients.

Figure 1

Kinetics of β-d-1,3-glucan and mannan in candidaemic patients and controls. Levels of (A) β-d-1,3-glucan (BDG) and (B) mannan were measured in 341 sera available from 43 patients with positive blood cultures. Sera were collected as a function of the number of weeks of hospitalisation in the ICU, with D0 as the day on which the first positive blood culture was taken. Levels of (C) BDG and (D) mannan were measured in 203 sera available from 67 control patients hospitalised on the same ward as a function of number of weeks of hospitalisation (control group). cGLC, concentration of BDG; cMan, concentration of mannan.

Figure 2

Kinetics of β-d-1,3-glucan, mannan and mannan antibody in five patients with candidaemia relapses. The dates of subsequent Candida isolations from blood are indicated by arrows. All patients, except Patient 4 with no detectable mannan, had values above the cutoff point for all tests, but with different kinetics. β-d-1,3-glucan (BDG) and mannan presented different kinetics of circulation in a single patient. The well-known transient nature of circulating mannan was observed and correlated with an inverse evolution of mannan antibody (Patients 1, 2, 3 and 5). Interestingly a sharp decrease in BDG was also observed over a short period of time (Patients 3, 4 and 5). cGLC/cBDG, concentration of BDG; cMn, concentration of mannan; Mn Ab, concentration of mannan antibody.

Figure 3

Receiver operating characteristic curves for β-d-1,3-glucan (plain line) and mannan (dotted line) for the period from day 7 before to day 7 after the first positive blood culture. Area under the curve (95% confidence interval) for β-d-1,3-glucan (BDG) = 0.71 (0.61 to 0.81) with P < 0.0001; threshold for the best sensitivity/specificity ratio, 353 (0.65 to 0.74). Area under the curve (95% confidence interval) for mannan = 0.63 (0.52 to 0.74) with P = 0.009; threshold for the best sensitivity/specificity ratio, 0.2 (0.36 to 0.94). When considering how the two tests could complement each other, mannan has an important contribution to specificity for low BDG values while higher BDG values have improved sensitivity and specificity. cGLC, concentration of BDG; cMn, concentration of mannan.

Figure 4

Biomarker-based algorithm proposed for managing preemptive therapy in ICU patients at high risk of developing invasive candidosis. IC, invasive candidosis.