We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
AUTOPSY-BASED REAPPRAISAL OF THE SERUM β-GLUCAN ASSAY FOR THE DIAGNOSIS OF INVASIVE FUNGAL INFECTION.
- Authors
Obayashi, Taminori; Funata, Nobuaki
- Abstract
PURPOSE To reexamine the efficacy of serum Β-glucan assay in the diagnosis of invasive fungal infection. METHODS A total of 456 autopsy records consecutive over the 6-year period from January, 2000 to December, 2005 were reviewed for the presence or absence of invasive fungal infection. The sensitivity of the assay was calculated based on those cases that were pathologically proved to be invasive fungal infections and had been tested for serum Β-glucan within 2 weeks before death. Likewise the sensitivity of blood culture was obtained for comparison. The specificity of the Β-glucan assay was determined on those cases that had no evidence of invasive fungal infection and had had a glucan test within 2 weeks before death. The cutoff was set at 30 pg/mL, because the upper limit of the normal range was 25 pg/mL. The positive and negative predictive values of the test were estimated against cutoffs at 30, 60, and 80 pg/mL based on a model population whose prevalence of invasive fungal infection was simulated to that of the present series. In addition, the concordance between serum Β-glucan assay and blood culture was studied by reviewing 'all the records of blood culture over the same 6-year period and collecting cases that grew fungi. Those who had been tested for serum Β-glucan within 2 weeks of blood sampling for culture were selected and the concordance was calculated. RESULTS Out of 456 autopsies, 54 (12%) had invasive fungal infections. Leukemias were the most frequent underlying disease (52%), followed by lynrphomas (13%) and AIDS (13%). The most frequent causative fungus was Aspergillus (70%). Out of the 54 cases with fungal infection, serum Β-glucan was measured within 2 weeks before death in 41 cases, and 39 of those had a concentration higher than the cutoff value. One of the two negative cases was infected with Zygomycetes, a fungus that does not produce Β-glucan. With this case excluded, the sensitivity of the serum Β-glucan assay was 97.5% (39/40). Blood culture was performed in 48 of the 54 cases, and four of them were positive for fungi. The sensitivity of blood culture was thus only 8.3%. Out of 402 cases without fungal infections, 63 was tested for serum Β-glucan within 2 weeks before death, and 54 of them showed the concentration lower than the cutoff. Thus, the specificity of Β-glucan assay was 85.7%. The positive and negative predictive values of the Β-glucan assay were 46.3% and 99.6% with cutoff at 30pg/mL; 70.9% and 98.3% at 60 pg/mL, and 87.0% and 97.4% at 80 pg/mL. During the same 6 years, there were 21 fungus-positive blood cultures that were preceded or followed by serum Β-glucan measurement within two weeks. The concentration of Β-glucan was above 30pg/mL in 17 cases. The concordance with blood culture, therefore, was 81% with the latter as a gold standard. CONCLUSION Serum Β-glucan assay is a highly sensitive and highly specific test for invasive fungal infection. With a cutoff set at 80 pg/mL, the test will rule in or rule out the diagnosis of invasive fungal infection with about 90% probability in hospitals where many patients with haematological malignancies and cancers are cared for
- Subjects
GLUCANS
- Publication
Mediators of Inflammation, 2007, Vol 2007, Issue 2, p9
- ISSN
0962-9351
- Publication type
Abstract