The control of tuberculosis (TB) in developing countries is hampered by the progression of drug-resistant TB, especially multi-drug resistance (MDR). MDR-TB strains are resistant to at least the first-line drugs isoniazid (INH) and rifampicin (RMP), which significantly reduces the therapeutic arsenal. The weighted rates of resistance in the world population are estimated among new cases at 17% for any resistance and 2.9% for MDR, and among previously treated cases at 35% for any resistance and 15.3% for MDR [1]. The World Health Organization (WHO) estimates that each year 440,000 new cases of MDR-TB are emerging globally, and 150,000 patients with MDR-TB die. More worrying is the emergence of extensively drug-resistant TB (XDR-TB) caused by MDR strains that also resist to fluoroquinolones and second-line anti-TB injectable drugs (amikacin, kanamycin or capreomycin). WHO estimates that 25,000 cases of XDR-TB emerge each year worldwide. Unfortunately, only a small fraction of patients infected by MDR or XDR-TB can access a screening by culture, drug-susceptibility testing and appropriate management [2]. The use of new rapid molecular tests for first-line drug resistance can dramatically reduce the delays to detect MDR-TB and adapt the treatment. The good performances of molecular line probe assays have led WHO to approve their use for rapid screening of patient at risk of MDR-TB [3].

In Lao PDR, the burden of TB remains high: it is the seventh leading cause of death in adults. In 2009, the prevalence of all forms of TB was estimated at 289/100 000, the incidence at 151/100 000 and the mortality at 24/100 000 [4]. The DOTS was implemented in 1995 by the National Tuberculosis Control Program (NTCP) [5]. The recommended treatment uses combinations of 5 drugs at fixed doses (S: streptomycin, R: rifampicin, H: isoniazid, Z: pyrazinamide, E: Ethambutol) corresponding to the following protocols: 2RHZE/6HE for new TB cases, 2RHZE/4RH for TB-HIV co-infections and 2RHZES/6RHZE for re-treatment. The therapeutic success rate of new smear positive cases exceeds 90% since 2005 and reached 93% in 2009. However, a retrospective cohort study of TB patients showed that the 85% survival rate was lower than the expected 91% from official statistics and that follow-up should be strengthened to improve patient compliance and allow contact tracing and detection of new cases [4, 6]. The implementation of the first mycobacterial cultures in 2009 has made feasible the anti-TB drug susceptibility testing [7]. The creation of a national reference laboratory for TB allowed the realization of a first national prevalence survey in 2010–2011. But more efforts should be implemented to achieve the Millennium Development Goal 6 of halving TB prevalence and reverse its incidence by 2015 [8].

In this context, this study aimed to provide a preliminary assessment of the rate of M. tuberculosis resistance to isoniazid (INH), rifampicin (RMP) and second-line drugs in Lao PDR using a molecular line probe assay.

The recent implementation of mycobacterial culture in Laos PDR had made possible the necessary monitoring of resistance to anti-TB drugs. This multicenter study aimed to provide preliminary data on the rate of resistance to first-line anti-TB, especially INH and RMP. Patient enrolled were outpatients, not referred from peripheral clinics, thus not being at higher risk of drug resistant. The results show an overall rate of resistance of approximately 10%, homogeneously distributed among the 3 sites. It reached 6.8% among new cases and 21.4% among previously treated cases. Despite a sampling not as robust as national survey, this study shows similar rates compared to neighboring countries: Cambodia, 10% and 16.6% in the 2001–2002 national survey [10], and Myanmar, 10% and 30.2% in the 2002 national survey [11]. However, it remains well below the rates reported in Thailand, 20.9% and 53.6% in 1996 [12], in Vietnam, 26.3% and 62.9% in 2001 [13] and in China, 27.9% and 60.3% in a meta-analysis covering the period 1993–2011 [14].

All the resistant isolates of this study showed mutations conferring resistance to INH. The only isolate resistant to RMP, also carrying mutations conferring resistance to second-line antibiotics, came out to be an XDR. Even though no strictly MDR strain was isolated during the study, these results show the existence of resistance mutations in TB isolates that may lead to compromise the effectiveness of treatment regimens recommended by the NTCP.

This study presents a number of limitations. The patient sampling was performed in three large hospitals which do not represent the whole country. A recruitment in remote district hospitals where patient follow-up and compliance to DOTS are less easy to perform [6] might have revealed higher levels of resistance. The HIV status of patients was unknown for 92% of patients, indicating that the recommendations for routine screening of HIV-TB co-infection had not been yet systematically applied. The exclusive use of a molecular test for rapid species identification of mycobacterial isolates and for detection of antibiotic resistance may also lead to underestimate INH resistance. This line probe assay is part of rapid methods recommended by WHO for detection of MDR strains as a substitute for conventional methods which require 4 to 6 weeks. It can even be performed directly from sputum samples found positive at microscopy. Rapid detection of resistance to RMP is a major issue because it can be considered a surrogate marker for MDR strains [15]. According to a meta-analysis of 14 comparative studies for RMP and 15 for INH, performances of the GenoType MTBDR assay are considered satisfactory, except for the detection of INH resistance whose sensitivity approaches 90% with the MTBDRplus assay [16]. Some strains phenotypically resistant to INH are found genetically susceptible, possibly because of mutations involving genes other than inhA and katG[17]. Similarly, the MTBDRsl assay showed an excellent specificity in detecting resistance to second-line drugs, but a variable sensitivity, especially for the detection of resistance to fluoroquinolones and ethambutol [18, 19]. The inclusion of new alleles is recommended in order to reduce the proportion of false negative results [20].

The eight resistant strains isolated in our study all showed a mutation conferring resistance to INH: 5 in katG, 2 in inhA and the remaining 1 in both katG and inhA. Interestingly, the three inhA promoter mutants were observed in new cases,. A study conducted during a 9-years period in San Francisco showed that strains with a katG S315T or inhA promoter mutation were more likely to spread than otherwise mutated strains [21]. However, mutations in the inhA promoter are more common among INH-monoresistant isolates than among MDR isolates [22]. It is in all cases a C15T mutation which is associated with a low level of resistance [23]. This mutation confers co-resistance to INH and ethionamide [24].

Of the 6 strains mutated in katG, 5 had the S315T mutation, currently considered as a risk factor for further development of MDR-TB [25]. The last one had a complete deletion of this gene, a mechanism rarely found associated with clinical INH-resistance [26, 27]. By sequencing, the distinction of three clones each carrying one peculiar mutation in the ahpC gene is consistent with the hypothesis that up regulation mutations in the ahpC promoter compensate for the loss of catalase-peroxidase activity encoded by the katG gene [28].

The only case of RMP-resistance was due to a mutation H526Y in the rpoB gene. With codons 531 and 516, this point mutation is one of the most frequently associated with resistance to RMP [29].

Four strains, two from new cases and two from previously treated cases, exhibited a C1402T mutation in the rrs gene, presumed to confer resistance to aminoglycosides and cyclic peptides: this mutation was found associated with high levels of resistance to capreomycin, but with low levels of resistance to kanamycin and amikacin [30]. Regarding the two new cases, this mutation strongly suggests a primary resistance whose origin is unclear, the only aminoglycoside antibiotic widely used in Lao PDR as first-line treatment until the 2000s being the streptomycin.

Despite the progress made by the NTCP to achieve the WHO objectives, implementation of the DOTS strategy in Lao PDR is limited by the remoteness of the target population to health centers, the difficult follow-up of patients under TB treatment and the lack of compliance. The recent development of mycobacterial culture and the use of line probe assays have allowed identifying the species of the M. tuberculosis complex and, for the first time, to detect resistances to anti-TB drugs. This preliminary investigation shows that resistance mutations, especially against INH, are present among the new cases, some of which may evolve towards MDR.

The isolation of an XDR strain from a patient enrolled in a provincial hospital is the result of chance. This strain acquired resistance as a result of successive treatments; XDR being the consequence of adding only 2 drugs never received to a previous treatment failure, instead of 5 as recommended by WHO. In Cambodia, of 101 MDR strains isolated between 2007 and 2009, only one was found XDR [31]. However, this must sound like a threat because the XDR strains have the ability to spread in the population by intra-community transmission [32] and justifies the strengthening and extension of the laboratory network for tuberculosis in the different provinces of Lao PDR. The use of line probe assays such as MTBDR tests should now be recommended to rapidly detect and treat without delay the MDR-TB cases.

The frequency of primary resistance to INH and identification of the first XDR strain strongly justify to strengthen and expand the screening of resistant TB in the provinces of Lao PDR.