Λοιμώξεις από mycobacterium spp. Ταυτοποίηση ειδών με βιοχημικές μεθόδους και με μεθόδους μοριακής βιολογίας

Abstract

Re-emergence of tuberculosis in combination with the appearance of multidrug-resistant strains in recent yearsintensifies the need for application of rapid methods for the identification of mycobacteria. Even though staining for acid-fast bacilli (AFB) to identify Mycobacterium tuberculosis complex (MTB) is usually performed in the first 24 h of specimen receipt, it is considered a method of low sensitivity. Although time consuming, the standard detection method remains the culture of specimens with optimum results obtained after application of both solid and liquid media.The use of nucleic acid amplification techniques provides a sensitive and specific approach for the identification of MTB directly in clinical specimens. The Cobas Amplicor polymerase chain reaction for MTB (CA PCR, Roche Diagnostic Systems, Inc., Branchburg, N.J., USA) is a system that combines specimen processing with automated amplification and detection, easily adopted by a clinical laboratory. In the present study, the sensitivity, specificity, positive and negative predictive values of CA PCR and culture results of AFB-positive and AFB-negative respiratory specimens obtained by different means (expectoration, bronchoscopy, expectoration after broncho­scopy, lavages and aspirations), as well as in samples from normally sterile body fluids, were evaluated. A total of 3414 specimens from 2365 patients (1 to 3 specimens/patient) with clinical suspicion of tuberculosis were processed during the study period (1999-2003) by the Microbiology Laboratory at the University Hospital of Patras. The respiratory specimens consisted of 684 sputa, 1473 bronchoalveolar lavages (BAL), 625 sputa expectorated after bronchoscopy (SAB), 296 tracheal aspirations (TA) and 189 pleural fluids. Furthermore,23 gastric aspirates and 124 samples ofsynovial, pericar­dial, peritoneal and cerebrospinal fluids (CSF) were also examined in the present study. All specimens were processed according to conventional procedures for identification of mycobacteria; cultures were performed by inoculation of sediments directly on Lowenstein-Jensen slants (L-J, bioMerieux, SA Lyon, France) and in BACTEC MYCO/F-Sputa and BACTEC MYCO/F LYTIC culture vials (Becton Dickinson). Statistical analysis was conducted using the SPSS v.12.0 software package for Windows (SPSS Inc.). Comparison of Z-N and PCR results were expressed by means of percent agreement and Kappa statistic. Highest sensitivity of CA PCR was observed among all positive AFB samples, whereas sputa collected after bronchoscopy were the most appropriate specimens, showing 70% sensitivity and 98.6% specificity, among the AFB-negative samples. The increase of M. tuberculosis infections is a global health problem in terms of both the disease and resistance to commonly used drugs. For the reason of continuing develop­ment of resistance and especially multidrug-resistance of M. tuberculosis (MDR), Microbiology Laboratories should provide reliable antibiotic suscep­tibility testing results in the minimum of time. In the present study, antimy­co­bacterial drug susceptibility testing (AST) was performed on 88 non-replicate M. tuberculosis clinical isolates, using the Mycobacterium Growth Indicator Tube (MGIT) System and Etest as compared to the method of proportion (MOP), in order to evaluate the potential application of these manual methods in the routine diagnostic laboratory. Obtained results among four antituberculous agents, isoniazid (INH), rifampin (RIF), ethambutol (EMB) and streptomycin (STR) were compared. Isolates were recovered from different patients and were identified at species level by PCR and hybridization. Resistance to INH was detected in 20.5%, 29.5% and 12.5% of the isolates, followed by STR resistance (19.3% 26.1% and 1.1%), RIF (9.1%, 4.5% and 5.7%) and EMB (2.3%, 11.4% and 2.3%) as showed by the MOP, MGIT and Etest, respe­ctively. Sensitivity of the manual MGIT ranged from 37.5% for RIF resistance to 100% for EMB, while sensitivity of Etest ranged from 5.9% for STR to 62.5% for RIF. In conclusion, whilesputum samples and BAL are the preferable clinical specimens for MTB recovery, sputa expectorated after bronchoscop are those showing the highest sensitivity by PCR, both among Z-N-positive and negative specimens. The combination of Z-N and CA PCR including internal control of collected after bronchoscopy from a patient with clinical signs of infectionundoubtedly contribute to the early diagnosis of tuberculosis. According to our data, the sensitivity of the manual MGIT is higher in testing resistant isolates compared to Etest, with the exception of rifampin. On the other hand, Etest shows higher specificity as well as higher positive predictive values and it may be used for testing rifampin resistance. However, for a routine Clinical Microbiology Laboratory, among the manual methods tested, the method of proportion remains the “gold standard” even though a longer incubation period is needed.