Tuberculosis (TB) is a bacterial disease caused by organisms of the Mycobacterium tuberculosis complex (MTC). The definitive diagnosis of TB depends on the isolation and identification of the etiologic agent. Clinical identification of mycobacteria remains difficult and timeconsuming. Cultures in specific media can result not only in M. tuberculosis (MTB) growth but also growth of nontuberculous mycobacteria (NTM). Therefore, the analysis of cultured colonies should be able to discriminate between MTB and NTM to confirm MTB specifically. Commercial diagnostic methods employ molecular biological tests to provide quick and specific tests for the identification of MTC, but false positive results cannot be excluded and these tests remain costly in terms of specialised equipment, labour and time (Wang 2006). MPT64, a 24 kDa secretory protein, is one of the major antigens from TB bacteria. Recently, MPT64 has been shown to differentiate the MTC from other bacterial species (Tomiyama et al. 1997, Abe et al. 1999, Hasegawa et al. 2002. Standard Diagnostics, Inc (SD) (Yongin, Korea) developed the SD BIOLINE TB Ag MPT64 RAPID ® test, which is a simple and rapid assay using a mouse monoclonal anti-MPT64 antibody that is able to discriminate between MTC and NTM by immunochromatography. The purpose of this study was to evaluate the utility of the SD BIOLINE TB Ag MPT64 RAPID ® test for the routine identification of mycobacteria in Madagascar.A total of 171 MTB and NTM strains were used for this study (Table I). Eleven samples of the H37Rv strain were used as standard controls. Clinical strains of MTB (88) and NTM (5) were isolated from clinical specimens and cultured on the Löwenstein Jensen (LJ) medium following decontamination and liquefaction procedures (Tison & Carbonelle 1972, Kent & Kubica 1985. Samples were inoculated in LJ medium and incubated at 37ºC for growth for several weeks. For Mycobacterium bovis (15 isolates) and NTM (14 isolates), strains from the mycobacteria unit collection were inoculated in 8 mL of Dubos medium and incubated at 37ºC for one week (for NTM) or three weeks (for M. bovis). Two hundred microlitres of each culture was inoculated in LJ medium and incubated at 37ºC. M. bovis strains (15) and NTM strains (3 of Mycobacterium chelonae, 2 of Mycobacterium avium, 2 of Mycobacterium intracellularae, 2 of Mycobacterium flavescens, 1 of Mycobacterium peregrinum, 1 of Mycobacterium vaccae, 1 of Mycobacterium aurum, 1 of Mycobacterium xenopi and 1 of Mycobacterium mucogenicum), after cultivation in Dubos media, were applied to the immunochromatography test slide directly without any manipulation.Mycobacteria strains were identified by their growth rate, colonial morphology, pigmentation, testing for urease, semi-quantitative catalase, heat-stable catalase, nitrate reduction and Tween 80 hydrolysis (David et al. 1989). In this study, 10 strains of M. bovis BCG (Pasteur) were used as a negative control for the MPT64 antigen as reported by Li et al. (1993). Other microorganism isolates (bacteria and fungi) ...

BackgroundTo improve the diagnosis of tuberculosis (TB), more rapid diagnostic techniques such as antibody detection based on immunochromatographic methods were developed. The objective of this study was to evaluate the performance of the SD Rapid TB kit for the diagnosis of active TB with serums from patients and close contact controls in Antananarivo, Madagascar.FindingsWe conducted a population-based case-control study. The sera of 60 confirmed TB patients and 60 healthy contacts paired for sex and age were tested. The controls were healthy contacts who were exposed to TB but had no clinical or radiological evidence of TB. The SD Rapid TB kit was used with serum samples according to the instructions of the manufacturer. Sensitivity, specificity, accuracy and predictive values were calculated using culture on Lövenstein-Jensen media as "gold standard".In this study, the sensitivity of the test was 55% and the specificity 90%. The positive predictive value and the negative predictive value were 84.61% and 66.66%, respectively, for diagnosis of pulmonary TB.ConclusionsThis study shows that the SD Rapid TB test is a simple and fast method. This test has a good specificity and could therefore help rule in TB if positive, but lacks adequate sensitivity.