ABSTRACT
Diagnosing pleural tuberculosis (plTB) might be difficult due to limited sensitivity of conventional microbiology tools. As M. tuberculosis (MTB)-specific T cells are recruited into pleural space in plTB, their detection may provide useful clinical information. To this aim, in addition to standard diagnostic tests, we used the QuantiFERON-TB Gold In-Tube (QFT-IT) test in blood and pleural effusion (PE) samples from 48 patients with clinical suspicion of plTB, 18 (37.5%) of whom had confirmed plTB. Four of them (22.2%) tested positive with a nucleic acid amplification test for MTB. The tuberculin skin test was positive in most confirmed plTB cases (88.9%). Positive QFT-IT tests were significantly more frequent in patients with confirmed plTB, as compared to patients with an alternative diagnosis, both in blood (77.7 vs 36.6%, p=0.006) and in PE samples (83.3% vs 46.6%, p=0.02). In addition, both blood and PE MTB-stimulated IFN-gamma levels were significantly higher in plTB patients (p=0.03 and p=0.0049 vs non-plTB, respectively). In blood samples, QFT-IT had 77.8% sensitivity and 63.3% specificity, resulting in 56.0% positive (PPV) and 82.6% negative (NPV) predictive values. On PE, QFT-IT sensitivity was 83.3% and specificity 53.3% (PPV 51.7% and NPV 84.2%). The optimal AUC-derived cut-off for MTB-stimulated pleural IFN-gamma level was 3.01 IU/mL (77.8% sensitivity, 80% specificity, PPV 68.4% and NPV 82.8%). These data suggest that QFT-IT might have a role in ruling out plTB in clinical practice.
Subject(s)
Interferon-gamma/immunology , Mycobacterium tuberculosis/immunology , Tuberculosis, Pleural/diagnosis , Adult , Aged , Aged, 80 and over , Enzyme-Linked Immunosorbent Assay , Enzyme-Linked Immunospot Assay , Female , Humans , Male , Middle Aged , Sensitivity and Specificity , Tuberculosis, Pleural/immunologyABSTRACT
The in vitro activity of 16 antimicrobial agents against 46 drug-resistant strains of Mycobacterium tuberculosis recently isolated from Italian patients was determined. As for first-line antituberculosis drugs, while isoniazid was ineffective against all the strains tested, resistance to streptomycin, rifampicin, pyrazinamide, and ethambutol was 80.4%, 71.7%, 39.1%, and 8.7%, respectively. Among second-line antituberculous drugs, resistance to ciprofloxacin, ofloxacin, and sparfloxacin and to amikacin and kanamycin was around 20%. About 10% of the strains were resistant to capreomycin and cycloserine and 4.3% were resistant to ethionamide; no strain was found to be resistant to thiacetazone, para-aminosalicylic acid, and viomycin. Although all strains displayed a rather continuous distribution of minimal inhibitory concentrations (MICs), a bimodal distribution was observed for rifampicin, amikacin, and kanamicin, with very high MIC values for resistant strains; relatively low MICs were found for fluoroquinolone-resistant strains. Among the small number of strains resistant to second-line agents, low resistant levels were observed. Restriction fragment length polymorphism analysis showed few strain clusters with resistance to first-line antituberculous drugs and aminoglycosides, fluoroquinolones, or both. Altogether, these results showed that second-line agents were still active against the isoniazid-resistant and multiply first-line resistant strains tested, with none or low resistance levels; these observations can be of importance for the treatment of multidrug-resistant tuberculosis in Italy.
