L. Lehtonen et al., DIAGNOSIS OF DISSEMINATED CANDIDIASIS BY MEASUREMENT OF URINE D-ARABINITOL L-ARABINITOL RATIO/, Journal of clinical microbiology, 34(9), 1996, pp. 2175-2179
Relative amounts of D-arabinitol (fungal origin) and L-arabinitol (par
t of normal human metabolism) in urine were determined by gas chromato
graphy and mass spectrometry from 61 hospitalized patients with hemato
logical malignancies. Seventeen neutropenic patients with acute leukem
ia (with 53 samples) had disseminated yeast infections and received em
piric antifungal therapy before confirmation of the diagnosis. Control
groups consisted of 22 hematologic patients (76 samples) with either
mucosal (n = 10) br urinary (n = 12) Candida colonization and 22 neutr
openic patients (34 samples) with no clinical or laboratory signs of i
nvasive yeast infection. Reference values were also obtained from 50 h
ealthy adults (50 Samples). The mean urine D-arabinitol/L-arabinitol r
atio +/- standard deviation (range) was 16.91 +/- 41.79 (1.41 to 254.7
5) in patients with disseminated infection, 2.73 +/- 2.48 (1.11 to 19.
00) in colonized hematologic patients, 2.12 +/- 0.84 (1.16 to 5.84) in
neutropenic controls, and 1.95 +/- 0.34 (0.97 to 3.44) in healthy adu
lts (P < 0.001 between patients with disseminated infection and all co
ntrol groups). The sensitivity and specificity of the assay for detect
ing disseminated yeast infection were, respectively, 88 and 91% per pa
tient (upper limit of normal, 4.00). Seventy-one percent of patients a
lready expressed elevated values at the onset of empiric antifungal th
erapy. The diagnosis of disseminated infection was confirmed on averag
e 21.7 days after the first elevation of the D-arabinitol/L-arabinitol
ratio. The method contributes to diagnosis of disseminated yeast infe
ction and helps in monitoring patients at risk, to support the initiat
ion of antifungal therapy at an early stage of the disease.