ABSTRACT
A polymerase chain reaction (PCR) assay was developed for the detection in clinical samples of mycobacteria belonging to the Mycobacterium tuberculosis complex. PCR products were detected with a simple and rapid colormetric method. With this method, 50 fg of M. tuberculosis DNA were detectable with the repetitive DNA-sequence-derived primers, corresponding to 10 genome equivalents. Detection of M. tuberculosis in 258 clinical samples by PCR was compared with detection by culture. PCR was positive for 56 of 57 culture-positive and Ziehl-Neelsen-staining-positive (ZN) samples, 11 of 18 culture-positive and ZN-negative samples. The presence of groEL DNA sequences was also investigated by PCR for all the specimens with the same revelation protocol. Three of the eight false-negative samples with the repetitive element-derived primers were found to contain groEL DNA sequences specific for the Mycobacterium genus. Among the 183 culture-negative samples, 30 were positive by PCR. When clinical data were known, the diagnosis of tuberculosis was established for the patients from whom those samples had been obtained. The results show that the rapid and simplified PCR assay described here is slightly more sensitive than culture and can be used in routine clinical practice.
Subject(s)
DNA, Bacterial/analysis , Mycobacterium tuberculosis/isolation & purification , Polymerase Chain Reaction/methods , Tuberculosis/diagnosis , Bacteriological Techniques , Cerebrospinal Fluid/microbiology , Colorimetry/methods , DNA Primers , DNA, Bacterial/genetics , Enzyme-Linked Immunosorbent Assay , Humans , Mycobacterium tuberculosis/genetics , Mycobacterium tuberculosis/growth & development , Oligonucleotide Probes , Reproducibility of Results , Sensitivity and Specificity , Sputum/microbiology , Urine/microbiologyABSTRACT
Five commercial antifungal susceptibility testing systems were studied for repeatability and reproducibility as well as concordance of results with the MICs for ten reference strains belonging to six different species. Repeatability was determined by testing each strain in triplicate on the same day, and reproducibility by repeating this triple determination on three different days. On the basis of 630 yeast-antifungal agent results for Mycototal and Mycostandard, 540 for Candifast, and 450 for ATB Fungus and Diff Test, repeatability was consistently equal to or greater than 95%. Reproducibility was 80.07% for Candifast and greater than 95% for the other systems. The concordance with the reference MICs was 51.65% for Candifast, 75.33% for ATB Fungus, 80.89% for Diff Test, 90.16% for Mycostandard and 90.32% for Mycototal. Although the performance of Diff Test and ATB Fungus was satisfactory, Mycototal and Mycostandard gave notably better results with imidazoles. Mycostandard, which is easier to use and includes tests for fluconazole and itraconazole, would seem to be potentially the most useful antifungal susceptibility test available at present.
