The Beijing genotype and drug resistant tuberculosis in the Aral Sea region of Central Asia

Tuberculosis (TB) remains one of the leading infectious killers worldwide, with an estimated 2 million deaths annually (1). In the year 2002 the number of incident TB cases was estimated at 8.8 million (2). Globally there is a 1.8% annual rise in new tuberculosis cases, with a 6% yearly increase in the former Soviet Union (1).

These data are increasingly accompanied by the phenomenon of drug-resistance, making successful treatment and control of the disease even more difficult. The third report on global surveillance for tuberculosis drug-resistance reveals alarming levels of MDR-TB (resistance at least to isoniazid and rifampicin) of up to 14% among new cases, with an estimated 300,000 new cases of MDR-TB globally per year [3]. Of particular concern are parts of Eastern Europe and Central Asia where tuberculosis patients are 10 times more likely to have MDR-TB than in the rest of the world [4].

The treatment of patients with MDR-TB is extremely difficult, expensive, and requires special treatment regimens and case management [5]. In addition, patients infected with MDR-TB may remain infectious for prolonged periods of times further accelerating the spread of MDR-TB. It is therefore important to understand factors contributing to the development of MDR-TB and the potential epidemiological impact of MDR-TB strains in order to develop effective control strategies.

Increasing tuberculosis incidence and the emergence of MDR-TB in the former Soviet Union have been mainly attributed to the deterioration of economic and social conditions as well as to the widespread breakdown of tuberculosis control systems since the late 1980s [6, 7]. The possibility that the pathogen itself is also contributing to this problem has been recently suggested by two studies from the Russian Federation, which found high proportions of a particular genotype of tuberculosis, namely the Beijing genotype, which was strongly associated with drug resistance [8, 9].

This notion has been further supported by recent studies that have provided evidence that the genetic heterogeneity of Mycobacterium tuberculosis complex isolates is greater than previously thought, and might influence the transmissibility and virulence of particular isolates [10‐13]. Strains of the Beijing genotype were first described in China and neighbouring countries in 1995 [14], and subsequently the occurrence of Beijing genotype strains has been documented in several parts of the world [8, 9, 14‐17]. The Beijing genotype has caused outbreaks of MDR-TB [18, 19], and some, but not all studies, indicate an association with drug resistance [16]. However, there is limited information available on both the occurrence and effects of the Beijing genotype, especially from areas with high tuberculosis incidence and high rates of MDR-TB. If these strains have an enhanced capacity to gain resistance, this will have serious consequences for the treatment of tuberculosis.

In a recent study, we found high levels of MDR-TB in the Aral Sea region in Uzbekistan and Turkmenistan, Central Asia [20]. A cross-sectional survey of more than 400 smear-positive tuberculosis patients, revealed levels of MDR-TB of 27% in Karakalpakstan (Uzbekistan) and 11% in Dashoguz (Turkmenistan). The DOTS strategy was introduced progressively into this region from 1998 and now covers a population of around 4 million people. High case notification rates for smear positive tuberculosis: 190/100,000/year in Karakalpakstan and 70/100,000/year in Dashoguz are reported from the DOTS programme.

In this study, we elucidate the importance of the Beijing genotype for tuberculosis in the Aral Sea region. Molecular typing of the isolates from the drug resistance survey was performed to determine the proportion of patients infected with Beijing genotype strains, and associations with drug resistance and other patient characteristics were analysed.

This study demonstrates that the Beijing genotype is a major cause of tuberculosis in this high incidence region of Central Asia. A strong association with drug resistance has been documented, independent from previous tuberculosis treatment. In addition, an association of Beijing genotype with clustering suggests that these strains are being transmitted throughout the community, possibly mediating the transmission of drug-resistant tuberculosis in this setting.

Since its description in 1995, the Beijing genotype of M. tuberculosis has elicited increasing attention. The prevalence of the Beijing genotype shows strong geographical variation from below 10% to more than 90% of the strains analysed [16]. High rates of Beijing genotype strains have been reported from some regions in Eastern Europe, such as Estonia, Azerbaijan, and Russia (Samara oblast, Archangel oblast, north-western region), while in Western European countries the prevalence of Beijing genotype is low [8, 9, 16, 26‐32]. Although a high prevalence of Beijing genotype strains appears to be confirmed for some regions of the world, direct comparison of data is difficult due to variations in strain identification methodology and different survey inclusion criteria utilised. Many of the studies performed so far have addressed subpopulations such as prison inmates or have provided only limited information on inclusion criteria [28, 29, 32]. These studies indicate high rates of Beijing genotype strains among these sub-populations, but definitive conclusions regarding prevalence in the general population are more difficult.

This study presents results on the prevalence of Beijing genotype among a representative cross-sectional sample in Central Asia. Since 50% of patients were found to be infected with Beijing genotype isolates, the data obtained confirm the importance of the Beijing genotype in this region and place this region in Central Asia among the Beijing genotype "high incidence" regions. Infection with Beijing genotype was not associated with particular subgroups of patients such as previous prison inmates or with a specific region suggesting that tuberculosis due to Beijing genotype is not restricted to specific parts of the population, but affects the general community.

This is of importance, since the Beijing genotype was additionally found to be significantly associated with drug-resistance and might be a driving force for the spread and emergence of MDR-TB. It is well known, that strains of the Beijing genotype have been involved in outbreaks of drug-resistant tuberculosis such as in the USA and in Russia [18, 19]. However, there was no consistent association with drug resistance among 12 studies evaluated in the review of Glynn and colleagues [16]. Only three studies found a strong association with resistance to particular drugs, while the other studies showed none or even a negative association.

In this investigation a strong association between the Beijing genotype and resistance to single drugs as well as with MDR-TB was observed. Beijing genotype was also associated with clustering when compared to non-Beijing strains, indicative of more recent transmission [33]. These data are in accordance with the results of the few recent studies so far performed in Eastern Europe: addressing tuberculosis in prison populations [28, 32], and the general population in the Archangel and Samara Oblasts and north-western regions of Russia and Estonia [8, 27, 29]. It is therefore likely that the Beijing genotype is a major cause of tuberculosis and particularly MDR-TB in wide areas of the former Soviet Union.

From our data, we cannot tell whether the Beijing genotype has emerged recently in the Aral Sea region or whether its long-term presence has been revealed. However, data from the Archangel oblast indicate that the Beijing genotype has been introduced in this region recently and has been strongly disseminating during the last few years [8]. Additionally, the Beijing genotype has also been reported to infect a rising number of patients on Gran Canaria Island, Spain, adding to the conclusion that strains of the Beijing genotype might have a selective advantage compared to other M. tuberculosis strains and spread more rapidly [34]. Therefore, in the former Soviet Union, the introduction of Beijing strains may have coincided with the deterioration of socio-economic conditions and the tuberculosis control system, and together this combination may contribute to the current MDR-TB epidemic.

The presence of multiple infection in 4% of the samples in this study confirms previous findings reporting mixed infections among small numbers of patients [35, 36]. The simultaneous infection of patients with a non-Beijing and a Beijing strain determined here, might point to specific properties of the Beijing genotype. In the majority of cases identified, the fingerprint pattern typical for the Beijing genotype was weak compared to the other pattern, indicating that an exogenous re-infection with a Beijing genotype strain might have occurred in these patients, which then out-competed the first isolate. A high rate of patients simultaneously infected with a Beijing and non-Beijing strain was also reported in a recent study in Cape Town, South Africa [37]. In this investigation, 19% of all patients were simultaneously infected with Beijing and non-Beijing strains. The possibility that resistant Beijing genotype strains can efficiently super-infect patients receiving regular tuberculosis treatment was further suggested by cases of exogenous re-infection with MDR-TB Beijing isolates described in previous studies [38, 39]. These findings suggest that susceptible or resistant Beijing genotype strains can potentially re-infect tuberculosis patients and in the case of resistant strains even whilst they are receiving regular tuberculosis therapy. This would be expected to result in a selective advantage for Beijing strains and would lead to a higher Beijing prevalence among drug resistance isolates.