Studies on laboratory diagnosis of fungal infection and antifungal susceptibility

Abstract

Invasive fungal infections (IFI) are serious complications in immunocompromised hosts. Neutropenic patients with haematological malignancies or after transplantation are at risk of developing deep Candida or Aspergillus infections. HIV-infected patients with T-cell defects are, on the other hand, predisposed to Cryptococcus neoformans meningitis and mucosal candidosis. Early diagnosis of IFI is of crucial importance since mortality is high even in treated patients.

The first aim was to assess the incidence of deep Candida infections and to improve their laboratory diagnosis. Candida colonisation and candidemia were retrospectively studied in patients with haematological diseases. Most patients were colonised by Candida and in the majority the same Candida species was isolated from at least two body sites. However, candidemia was diagnosed in only 3 (1.1%) patients. Possible causes of this low incidence were fluconazole prophylaxis, empirical amphotericin B treatment, and restrictive use of antibiotics.

A nested polymerase chain reaction (PCR) was developed for diagnosis of deep Candida infections. Candida albicans DNA was detected in serum and organ samples from experimentally infected mice as well as in serum from patients with candidemia or invasive Candida infection.

Serum and urine specimens, prospectively collected from transplant recipients with clinically suspected IFI before, during, and after empirical antifungal treatment, were analysed by PCR for detection of C. albicans DNA in serum and by measurement of D/L- arabinitol ratio in urine. Serum was also tested for concentration of (1->3)-ß-glucan and commercial Candida antigens. Invasive fungal infection (IFI) was verified in 4 candidiasis and 1 aspergillosis patients (13%). Positive non-culture assay was recorded in 3 of 5 IFI patients. Of the empirically treated patients 36% had >2 positive non-culture assays.

The second aim was to investigate animal susceptibility of fungi. Fluconazole resistance (MIC >32.0 mg/l) was found in 7 (14%) of the oral C. albicans strains isolated from HIV-infected patients. Two of the resistant strains showed cross-resistance to other azoles. A correlation between reduced susceptibility to fluconazole, low CD4+ cell count, and time interval from the first fluconazole treatment, indirectly reflecting the total fluconazole exposure, was observed.

Equal fluconazole MIC90 (8 mg/l) was obtained by macrodilution, microdilution, and E- test methods for C. neoformans strains, isolated from cerebrospinal fluid. In comparison with the reference NCCLS macrodilution method, good agreement was obtained by both microdilution and E-test techniques. Both amphotericin B and itraconazole IRC90 was 1 mg/l for Aspergillus strains isolated from lower respiratory tract. Acquired itraconazole resistance (MIC>32 mg/l) was found in A. fumigatus strains isolated from 3 patients during long term itraconazole therapy.

Conclusions: Combination of several non-culture assays is needed for laboratory diagnosis of IFI in high risk patients. No single test is sufficient for diagnosis. Antifungal susceptibility testing by standardised methods is important since resistance to azoles may develop during prolonged treatment.