A total of 221 isolates of M. tuberculosis were sampled from hospitals and the general population in the northern plain of Vietnam, one of the most populated region of the country. Genotypic composition and diversity were characterized, and we investigated how they are affected by sampling (hospital vs. general population), correcting for potential confounding effects (location, age and gender of the patients). Spoligotyping and 12 MIRU-VNTR typing were used as first line. Then 15 MIRU-VNTR standard set was used, making 21 MIRU-VNTR typing for the clustered isolates. Result showed that 8 lineages and 13 sub-lineages were circulating in the region. The most predominant lineages were Beijing (38.5%) and EAI (38.5%). Others appeared with small proportions H (1.4%), LAM (1.8%), T (8.1%), X (0.9%), MANU (2.3%), and Zero (0.4%). Higher clustering rate was found in the hospital samples (17.9% in urban and 19.2% in rural areas) compared to the population ones (0%). The typical Vietnamese EAI4-VNM sub-lineage of EAI lineage accounted for 67% of EAI strains and was associated with older ages. Beijing genotypes were associated with younger, urban population and were characterized by high clustering rates. These characteristics strongly suggest that Beijing strains are invading the population, replacing the local EAI-VNM4, thus predicting a more serious tuberculosis situation in the future in the absence of more effective control strategies.

BackgroundMycobacterium tuberculosis, the tuberculosis (TB) pathogen, despite a low level of genetic diversity, has revealed a high variety of biological and epidemiological characteristics linked to their lineages, such as transmissibility, fitness and propensity to acquire drug resistance. This has important implications for the epidemiology of TB. We conducted this first countrywide cross-sectional study to identify the prevalent M. tuberculosis lineages and to assess their epidemiological associations and their relation to drug resistance. The study was conducted among isolates acquired in reference hospitals across Vietnam. Isolates with drug susceptibility testing profiles were identified for their lineages by spoligotyping. Logistic regression was used to investigate the association of M. tuberculosis lineages with location, age and sex of the patients and drug resistance levels.ResultsResults showed that the most prevalent lineage was Beijing (55.4 %), followed by EAI (27.5 %), T (6.4 %), LAM (1.3 %), Haarlem (1 %) and Zero type (0.3 %). The proportion of Beijing isolates in the North (70.4 %) and the South (68 %) was higher than in the Centre (28 %) (OR = 1.7 [95 % CI: 1.4–2.0], p < 0.0001), whereas the proportion of EAI isolates in the North (7.1 %) and the South (17 %) was much lower compared with the Centre (59 %) (OR = 0.5 [95 % CI: 0.4–0.6], p < 0.0001). Overall, Beijing isolates were the most likely to be drug-resistant and EAI isolates were the least likely to be drug-resistant, except in the South of Vietnam where EAI is also highly drug-resistant. The proportion of Beijing isolates was significantly higher (p < 0.01), and the proportion of EAI isolates was significantly lower (p < 0.05) in younger patients. The proportion of drug-resistance was higher in isolates collected from male patients and from patients in the middle age groups.ConclusionsThe findings suggest ongoing replacement of EAI lineage, which is mainly more drug-susceptible with highly drug-resistant Beijing lineage in all studied regions of Vietnam. Male patients of working ages should be the focus for better control to prevent the emergence of drug-resistant TB.Electronic supplementary materialThe online version of this article (doi:10.1186/s12866-016-0784-6) contains supplementary material, which is available to authorized users.

No abstract

Multidrug-resistant tuberculosis is a major issue worldwide; however, accessibility to drug susceptibility testing (DST) is still limited in developing countries, owing to high costs and complexity. We developed a proportion method on 12-well microplates for DST. The assay reduced the time to results to <12 days and <10 days when bacterial growth was checked with the naked eye or a microscope, respectively. Comparison with the Canetti-Grosset method showed that the results of the two assays almost overlapped (kappa index 0.98 (95% CI 0.91-1.00) for isoniazid, rifampicin, streptomycin; and kappa index 0.92 (95% CI 0.85-0.99) for ethambutol). The sequencing of genes involved in drug resistance showed similar level of phenotype-genotype agreement between techniques. Finally, measurement of the MICs of rifampicin and ethambutol suggests that the currently used critical ethambutol concentration should be revised, and that the current molecular drug susceptibility tests for rifampicin need to be re-evaluated, as in vitro rifampicin-sensitive isolates could harbour drug resistance-associated mutation(s).