Prospective Study and Analytical Performance of Serum (1->3)-β-D-glucan in Pediatric Patients

Background: Invasive fungal infections (IFIs) are important causes of morbidity and mortality among immunocompromised patients (pts). Culture and histopathology are considered the gold standard for laboratory diagnosis. The fungal cell wall biomarker, (1→3)-β-D-glucan (BDG), improves diagnosis of invasive candidiasis in adults; there is very limited data available in pediatric pts

Methods: Between 2012 and 2015, pediatric pts (birth-21 yrs), admitted to New York Presbyterian Hospital/Weill-Cornell Medicine with confirmed positive fungal culture from a sterile site, were prospectively enrolled (Case Group). Low risk healthy pediatric pts were enrolled as a Control Group. BDG was obtained from Cases up to twice a week, preferably over 4 weeks, from time of diagnosis and once from Controls. Pts were classified by age cohorts, including preterms. Baseline (at diagnosis), maximum and median levels of BDG were determined. Demographics, predisposing conditions, immunosuppression, possible contaminants, site of isolation, type of yeast, antifungal prophylaxis, and therapy were recorded

Results: 27 cases and 100 controls were enrolled. Candida (C.) parapsilosis was isolated in 8/27 (29.6%) cases, followed by C. albicans 6/27 (27%); there were 2/27 (7.4%) Trichosporon asahii. There was no statistical difference of BDG levels at diagnosis, maximum, and median by age group, type of yeasts, presence of neutropenia, corticosteroids, immunosuppression, and antifungal prophylaxis. Median BDG level differed for Candida (n=22) 371.97 pg/mL (range (R) 49.31, 26637.72) vs non-Candida yeasts (n=4) 64.6 pg/mL (R 35.7, 109.5), p=0.0371. There was no statistical difference among all control age cohorts (p=0.94) for median BDG level: 41.7 pg/mL (preterms); 65.4 pg/mL (birth-12 months); 55.9 pg/mL (1-3 yrs); 50.8 pg/mL (4-6 yrs), 50.4 pg/mL (7-12 yrs) and 77.4 pg/mL (13-21 yrs). Median BDG levels of Cases at Diagnosis were significantly elevated in comparison to those of Controls, 371.9 pg/mL (R 24.5, 17189.4) vs 56.5 pg/mL (R 2.0,1556.0) respectively, p=<0.0001). With 80 pg/mL cutoff, sensitivity, specificity, and NPV of BDG at Diagnosis were 77, 67, and 91.7% respectively; while with 100 pg/mL, specificity increased to 76% and NPV to 92.6%

Conclusion: An 80 to 100 pg/mL breakpoint has sufficient NPV to exclude most IFIs in pediatric pts. BDG levels of low risk children are similar to those of adults.