Identification of members of the Mycobacterium tuberculosis complex and the M. avium-M. intracelludare complex (MAC) directly from primary BACTEC cultures was evaluated by using acridinium-ester-labeled DNA probes (AccuProbe; GenProbe, Inc., San Diego, Calif.). In preliminary experiments, blood present in samples was found to interfere with the assay because of nonspecific chemiluminescence, which was measured in relative light units (RLUs). There was a direct relationship between the age of

more »

... culture and the number of nonspecific RLUs. A protocol using 1% sodium dodecyl sulfate-5 mM EDTA to treat BACTEC broth cultures which, with specimens containing blood, gave on the average a ninefold reduction in nonspecific chemiluminescence was developed. By using this treatment protocol, 120 specimens were tested directly from BACTEC broth cultures with an AccuProbe for the M. tuberculosis complex and/or the MAC. In order to establish the background of the specimen, the patient sample was assayed without probe. The criteria for the inclusion of BACTEC cultures in the evaluation were a growth index of 2100 and a positive smear for acid-fast bacilli directly from the BACTEC broth. For the 120 cultures tested, if a hybridization result of 230,000 RLUs was considered positive, the sensitivities for detecting the M. tuberculosis complex and the MAC were 47 and 90%o, respectively, with a specificity of 100% for both. However, if a ratio of the RLUs obtained with the MAC or the M. tuberculosis complex probe to those obtained with the specimen background of >20 was considered positive, this gave 77% sensitivity and 100o specificity for BACTEC cultures containing M. tuberculosis complex isolates and 96% sensitivity and 100% specificity for those growing MAC isolates. DNA probes have aided greatly in the rapid identification and detection of Mycobacterium spp. from cultures of clinical specimens. The first generation of probes to be used on colonies isolated on solid media utilized an isotopic label (2). In a previous report, we used isotopically labeled probes to directly identify Mycobacterium spp. from BACTEC cultures (12). A second generation of nonisotopic probes, the SNAP system (Syngene, Inc., San Diego, Calif.) which utilizes DNA probes labeled with horseradish peroxidase (10), and the AccuProbe system, which employs acridinium ester as the probe label (5), have been introduced. With this latter system, the hybridization reaction is detected by chemiluminescence upon hydrolysis of the label by H202 and NaOH. In addition to the advantages that come with the substitution of the isotopic label, the second generation of assays provides both probes for Mycobacterium avium and M. intracellulare as well as a combined M. avium-M. intracellulare complex (MAC) probe, whereas the first-generation system had only separate probes for M. avium and M. intracellulare. In many settings in which several cultures are obtained from a patient and/or serial cultures are used to monitor antimicrobic therapy, it is not necessary to distinguish between M. avium and M. intracellulare with every specimen, and it is in these circumstances that a combined MAC probe offers time and cost benefits. However, specific M. avium and M. intracellulare probes are also useful in initially establishing the species isolated from a patient so * Corresponding author. that information on treatment protocols can be gathered and, in some cases, antimicrobial therapy can be tailored to the species (7). The AccuProbes have been reported to be both specific and sensitive for identifying Mycobacterium spp. (5). In comparing the AccuProbe system for identifying colonies from solid media with the earlier, isotopically labeled version of these probes, Goto et al. (5) reported 100% sensitivity for both the M. tuberculosis complex probes and the MAC probe. In this study, we evaluated the AccuProbe system for its ability to identify the M. tuberculosis complex and the MAC directly from BAC-TEC broth cultures. We have found that blood contributes to false-positive results with the Accu-Probe system, and we have developed a method to reduce this problem. MATERIALS AND METHODS Specimens. Specimens included in the study were those sent to the Medical Microbiology Laboratory at the University of California Irvine Medical Center with requests for mycobacterial cultures. Upon their receipt in the laboratory, the specimens were processed for culture within 24 h. Blood specimens were received in Isolator tubes (Wampole Laboratories, Cranbury, N.J.) which were centrifuged at 3,000 x g for 30 min, and the pellets were resuspended according to the instructions of the manufacturer. Specimens, including blood, that did not require further decontamination or digestion were inoculated directly into BACTEC 12B bottles (Becton Dickinson) as well as onto conventional media,