The World Health Organisation (WHO) estimates that if the current tools for the treatment and prevention of TB are not improved, nearly 1 billion additional people will become infected by M. tuberculosis between the 2000 and 2020. Among these, 200 million will develop the active disease and 35 million will die from TB. By the end of 2004, 199 (94%) of 211 countries notified 4.4 million new and relapsed cases, of which 1.9 million (44%) were new sputum smear-positive. Among these notifications, 3.7 million were from DOTS (Directly Observed Treatment Short course) areas. The Region of the Americas notified 235,187 cases in 2004 1. Venezuela covers an area of 916,495 km² and has a population of around 25 million inhabitants, and has shown a marked decrease in the incidence of TB during the last years. The TB prevalence rate in 1939 was 98.6 per 100,000 inhabitants for mortality and 111 per 100,000 inhabitants for morbidity 2. Since then, the incidence has decreased dramatically to reach 6734 cases (26 per 100,000 inhabitants), with most being confined within the 15 to 45 year age group. Nevertheless, in indigenous populations, TB incidence has remained very high (137.4 per 100,000 inhabitants) 3. In 1982, the National Tuberculosis Control Program of Venezuela (NTP) implemented the directly observed treatment and since 1994 has used a standardised six-month treatment for all new TB cases with or without culture confirmation 4.

Venezuela carried out the first survey of anti-tuberculosis drug resistance during 1998–1999 as part of the Global Project of Anti-tuberculosis Drug Resistance Surveillance, coordinated by the International Union Against Tuberculosis and Lung Disease (IUATLD) and the WHO. The survey enrolled 873 patients with smear-positive TB, and living in 23 states from Venezuela 45

The molecular characterisation of the isolates collected during this study was not carried out and there is no information on the major circulating clades of M. tuberculosis. The mutations involved in drug resistance and the question of whether the Beijing family of the M. tuberculosis complex had spread into Venezuela have not been studied until now. Molecular epidemiological studies have showed the widespread of this family and its association with drug-resistance 678.

In recent years, the genetic typing of M. tuberculosis complex (MTC) strains has been widely used to support conventional epidemiological investigations of TB outbreaks and as a tool for studying transmission dynamics. The analysis of restriction fragment length polymorphisms (RFLP), which is based on the variability of the chromosomal location and copy number of the insertion sequence IS6110, is considered the gold standard for the genotyping of MTC 910. However, this technique is expensive for developing countries, is cumbersome and requires a large amount of highly purified DNA, and is limited for strains with fewer than six IS6110 insertions or when rapid results are needed. Spoligotyping has emerged as a robust, easy, cheap and reliable supplementary tool for the molecular epidemiological study of TB. The establishment of an international database of spoligotypes provides information on the overall diversity of strain patterns. The combination of spoligotyping with bioinformatics has been proposed as a potential tool for defining major circulating clades of tuberculosis bacilli and to analyse the complexity of global TB transmission 11121314. Spoligotyping has also been used to track TB epidemics, detect new outbreaks, define high-risk populations and to help to verify suspected false-positive cultures due to laboratory contamination and to confirm nosocomial transmission. In conjunction with other genotyping assays, it has also been used to investigate the evolutionary genetics of the tubercle bacilli 15161718.

This present investigation aims to provide an initial idea of the genetic biodiversity of M. tuberculosis in Venezuela. We used spoligotyping as a first-line discriminatory test in 670 strains isolated from the 23 surveyed states during the first anti-tuberculosis drug-resistance survey. RFLP analysis of IS6110 and characterisation of the katG and rpoB genes were carried out on 14 of the 18 MDR strains included in this survey.

In 1994, the WHO and IUATLD launched a Global Project on Anti-Tuberculosis Drug resistance surveillance to assess the global extent of drug resistance. During 1998–1999, the NTP in Venezuela carried out the first national drug resistance survey through out the entire country for a period of nine months. The sampling method was the proportionate cluster method and the total of 873 patients were tested. The Venezuelan NTP has a long history of tuberculosis control and the results of the survey showed a very low prevalence (0.5%) of MDR-TB in never treated patients, similar to those reported in Canada, Cuba, Uruguay, and Chile, which are countries with effective NTCPs 45. However, the prevalence of resistance to at least one drug among the previously treated cases was 30.8% and MDR-TB was 13.5%.

We used spoligotyping to molecularly characterise 670 (77%) of the 873 isolates, including isolates of the 23 states surveyed. It was not possible to analyse all isolates by RFLP analysis IS6110. Nevertheless, as the molecular epidemiology of MDR M. tuberculosis in Venezuela is not know, we investigated the most of MDR isolates detected in the survey using spoligotyping, RFLP analysis of IS6110 and point mutations of the katG and rpoB genes.

Comparison of our spoligo-patterns with the international spoligotyping database revealed that 74 SITs had already been identified. We describe 20 new SITs specific to Venezuela that were found in 75% of the 23 provinces surveyed and show that these types are extensively dispersed. The six most prevalent spoligotypes (SITs 17, 42, 93, 53, 20 and 605) included the half of all the isolates. Among the clades of M. tuberculosis complex described as harbouring specific spoligotypes, 89% of the isolates studied belonged to three major genotypic families (LAM, T and Haarlem). The majority of them belonged primarily to the LAM family, which has a high prevalence in Latin-America, the Caribbean and the Mediterranean region. This family recently was reclassified in 12 sub-clades 21 and in this study we identified SITs corresponding to the sub-clades LAM-2, 9, 5, 1, 4, 3 and 6. However, the genetic network of spoligotypes showed that the LAM-2, LAM-9, and LAM-5 do indeed constitute the backbone of the genetic TB population-structure in Venezuela. A more detailed picture of this network would require further investigations using complementary genetic markers to better identity the clinical isolates within the LAM superfamily of M. tuberculosis, which is highly prevalent in Venezuela.

The "ill-defined" T genetic family encountered in different regions of the world 20 was the second most common family. Within this family we found SITs that belonged to the sub-clades T-1, 2, 3, 4 and 5 21. We also found some isolates belonging to the Haarlem family, sub-clades H-1 and H-3. This family has a European origin and has been described in the Caribbean and Central America, suggesting that it is remnant of the European colonisation 2324. The Beijing/W genotype was not detected in this study.

Our finding of seven new isolates belonging to SIT 333 (LAM-5), which has been reported in Guadeloupe, French Guiana, Haiti and Venezuela, suggest a common origin. SIT 605 had been reported in the international database as only being found in Venezuela and the USA, and our study adds other 18 clinical isolates, to this SIT, from patients living in seven Venezuelan provinces. Further studies are needed for a better understanding of the distribution of these spoligopatterns.

When the Venezuelan isolate spoligopatterns were added to the international database, eight new inter-regional clusters were created by matches with identical patterns from USA, Brazil, France, Latvia, Australia and The Netherlands. Curiously, one of these new inter-regional clusters (ST 1717), which included an Australian isolate, shared a pattern with one of our Venezuelan isolates belonging to an indigenous patient from the Delta Amacuro state.

We characterised the katG and rpoB genes in 78% (14/18) of the MDR strains isolated in the survey. We observed that 64% of the isolates had a mutation in the rpoB gene, whereas other studies have reported mutations in the 81-bp core region in more 90% of MDR strains 252627. It has been reported that a mutation outside of the 81-bp rpoB region associated with RIF resistance, and the heteroresistance caused by a mixture of mycobacterial subpopulations with different susceptibilities to RIF may influence the sensitivity of the molecular tests for detecting resistance 28. In our case, the rifampicin-resistance of these strains had been confirmed by two phenotypic methods, however the mutation in the rpoB gene was no detected in five of the fourteen strains analysed. Nevertheless, a new test to confirm the resistance of those strains could not be carried out because the cultures of the strain genotyped were not available. The Ser531Leu and Asp516Val mutations detected in this study are among the most common observed worldwide 293031.

Mutations in the katG gene have been reported between 20 and 80% of INH resistant M. tuberculosis strains. The mutation Ser-Thr at codon 315 of katG is the most common 32. In our study, 43% of the isolates showed this mutation, consistent with other studies 3334. Other mutations in of this gene and other gene regions, such as the inhA locus, the kasA, ndh, oxyR-ahpC intergenic region, have been implicated in INH resistance. However, we did not evaluate these mutations.

The characterisation of rpoB could discriminate between the three clusters grouped by RFLP analysis of IS6110 and by spoligotyping, as only one showed identical mutations suggesting a recent transmission of MDR tuberculosis 31. This cluster contained isolates recovered from the same province. The other two clusters had different resistance-associated mutations, indicating that these isolates had acquired mutations independently, and were probably transmitted before the acquisition of drug resistance.

There is limited information regarding the distribution of M. tuberculosis complex genotypes as well as few molecular studies on MDR strains have been carried out in the most Latin American countries. A recent study describing the genetic diversity of M. tuberculosis complex and the world-wide distribution of spoligopatterns 21 showed only a limited number of registered strains, particularly in regions outside the immediate vicinity of Venezuela. Greater financial means are needed to implement an adequate infrastructure and to develop molecular typing techniques for the national reference laboratory at NTP, which will be an important tool for the control programmes in these countries.

Although Venezuela has had, a well functioning NTP during many years, the proportion of immigrants from neighbouring countries with a high incidence of TB and MDR strains, the reduction of TB control activities, and a similar decrease in the locating and treatment of cases among indigenous populations may worsen the current TB situation in the coming years. Other national studies incorporating other molecular epidemiology methods are needed to evaluate the actual situation of MDR in Venezuela.

As it is important to clarify the molecular epidemiological characteristics for effective tuberculosis control, we used spoligotyping as a first-line discriminatory test. This is the first genetic biodiversity study of M. tuberculosis in Venezuela and our results should be considered a starting point for the creation of a national genotype database to support the National Tuberculosis Control Program in Venezuela.

LA conceived the study, carried out the molecular genetic studies, analysed the data and drafted the manuscript; RA, ME, AG and AVdeS participated in the design and carried out the first survey of anti-tuberculosis drug-resistance, analysed the data, and provided the clinical isolates for molecular study. CR and JR carried out mycobacteriological diagnostics, isolation of clinical isolates, identification, drug susceptibility tests and provided information about the clinical isolates; CL participated in the genotyping studies; AC assisted with data entry, conducted the bioinformatics programming and statistical analysis; MAL and MJR provided technical help in the conservation of the strains; CS and NR carried out the phylogeny reconstruction studies, participated in the identification and designation of the SITs and also helped draft the manuscript; CM participated in the design of the study, and provided critical comments for the manuscript; SS conceived the study, and participated in its design, coordination of the investigation, and helped to draft the manuscript for consideration for publication. All authors contributed to the study, read and approved the final manuscript.