Molecular analysis of drug-resistant Mycobacterium tuberculosis isolates collected in central Poland


Corresponding author and reprint requests: Z. Zwolska, Department of Microbiology, National Tuberculosis and Lung Diseases Research Institute, Płocka 26, 01-138 Warsaw, Poland


In total, 48 Mycobacterium tuberculosis clinical isolates from 48 patients with drug-resistant pulmonary tuberculosis (TB) in central Poland during 2004 were analysed by spoligotyping and IS6110-Mtb1/Mtb2 PCR. Twelve (25%) isolates were clustered by spoligotyping in combination with IS6110-Mtb1/Mtb2 PCR. Patients from whom these isolates were obtained were assumed to have developed TB as a result of recent transmission. Spoligotyping, used alone, overestimated the number of clustered isolates. However, combined use of spoligotyping and IS6110-Mtb1/Mtb2 PCR was an efficient approach for revealing clonal relatedness among M. tuberculosis isolates.

This report describes the molecular characterisation of drug-resistant Mycobacterium tuberculosis isolates recovered from tuberculosis (TB) patients in central Poland during 2004. This was achieved by using a combination of spoligotyping [1] and IS6110-Mtb1/Mtb2 PCR [2]. The phenotypic characteristics of the isolates were also determined. The study included 48 M. tuberculosis isolates from 48 unrelated adult TB patients (age range, 18–85 years; median age, 48 years; 38 men and ten women) from central Poland. The population examined represented 7% of all bacteriologically confirmed cases of TB notified in this region during 2004, and 20% of the total number of drug-resistant pulmonary cases of TB recorded in Poland during 2004.

Primary isolation was performed using Löwenstein–Jensen medium and the BACTEC  460-TB system (Becton Dickinson, Franklin Lakes, NJ, USA), with species identification based on niacin tests, the use of gene probes (Gen-Probe, San Diego, CA, USA) and HPLC analysis of mycolic acids. Drug susceptibility testing used the 1% proportion method in Löwenstein–Jensen medium. The drugs tested (and their critical concentrations) were: isoniazid (0.2 mg/L), rifampicin (40 mg/L), streptomycin (4 mg/L), and ethambutol (2 mg/L) [3]. Bacterial DNA was prepared by the cetyl-trimethyl-ammonium bromide method [4]. Spoligotyping was performed using a commercially available kit (Isogen Bioscience BV, Maarssen, The Netherlands) according to the manufacturer’s instructions and as described previously [1]. IS6110-Mtb1/Mtb2 PCR was also performed as described previously [2]. Isolates with indistinguishable spoligotypes and IS6110-Mtb1/Mtb2 patterns were assumed to be clonally related and to be linked by recent transmission events.

Drug susceptibility testing revealed that 38 (79%) isolates were resistant to only one drug. Isoniazid and streptomycin mono-resistance were represented almost equally, accounting for 33% and 37% of the total isolates, respectively. Nearly two-thirds of the isolates resistant to a single drug were from patients with no history of treatment. Eight (17%) isolates were multidrug-resistant (MDR), of which six were from previously treated patients.

In total, 26 distinct spoligotypes were identified among the 48 M. tuberculosis isolates tested, with 19 (40%) isolates represented by a unique pattern, and 29 (60%) isolates grouped in seven clusters, each containing two to eight isolates. Of the 19 unique profiles, 13 (27%) belonged to a previously described ‘shared type’ (ST) according to the worldwide spoligotype database (SpolDB4), while six did not match any existing spoligotype and were thus defined as ‘orphans’. Among the seven clusters recognised, one was not found in SpolDB4. The most frequent spoligotypes observed were ST53 and ST47, both representing 15 (31%) of the isolates studied. Also, two isolates of the Beijing genotype (ST1) were identified.

IS6110-Mtb1/Mtb2 PCR typing revealed that 17 (35%) isolates included in spoligotype clusters produced unique banding patterns. Only two spoligotype clusters, comprising two and four isolates, respectively, contained isolates that were identical according to IS6110-Mtb1/Mtb2 PCR. The 19 isolates with unique spoligotypes also generated unique IS6110-Mtb1/Mtb2 PCR patterns. Clustering results obtained according to both genotyping methods are summarised in Table 1. Overall, the combination of spoligotyping and IS6110-Mtb1/Mtb2 PCR assigned 12 (25%) of the 48 isolates to four different clusters, containing isolates from four, three, three and two patients, respectively (Table 2).

Table 1.   Summary of clustering results obtained according to the two genotyping methods used in this study
MethodNo. of clustered isolates (%)No. of clustersSize of cluster (no. of isolates)
Spoligotyping29 (60)72–8
IS6110-Mtb1/Mtb2 PCR12 (25)42–4
Table 2.   General characteristics of tuberculosis cases linked by recent transmission on the basis of clustering results
 ClusterPatient no.aType of resistancebDrug resistance profile
SpoligotypingIS6110-Mtb1/Mtb2 PCR
  1. aCorresponds to the isolate number.

  2. bP, primary; A, acquired.

  3. cSpoligotype not found in the worldwide spoligotype database.

  4. INH, isoniazid; SM, streptomycin; RMP, rifampicin.


The two most prevalent spoligotypes (ST53 and ST47) were responsible for almost one-third of the TB cases studied. These spoligotypes, along with ST50 and ST1051, were also the top-ranked genotypes among drug-resistant M. tuberculosis isolates in Poland during 2000 [5], and, except for ST1051, are among the most frequent worldwide [6]. In addition to the 19 spoligotypes defined previously, seven new types were identified, of which six were orphans and one, comprising four isolates, was a new ST. The presence of these new and unique spoligotypes, especially the one comprising a cluster, suggests that they are specific to the study setting. Overall, ten (21%) of the cases were caused by M. tuberculosis isolates whose spoligotypes had not been described previously.

Two isolates belonging to the Beijing family (ST1) were identified. This genotype has been associated frequently with drug resistance and TB outbreaks worldwide [7,8], and Beijing strains are very common in Poland’s eastern neighbours, Estonia and Russia, with prevalences of 29% and 44.5%, respectively [9,10]. Beijing strains have been described previously in Poland [5], which suggests a continuing presence of the Beijing genotype in the region; nevertheless, it was possible to discriminate between the two Beijing isolates in this study, which left them outside any chain of recent transmission.

Although spoligotyping is a convenient and rapid method for preliminary screening, it overestimates the number of epidemiologically linked cases [11,12]. Consequently, spoligotyping has been proposed as a first-line test in a two-step strategy for M. tuberculosis typing, to be followed by another PCR-based fingerprinting method with a higher discrimination level [11]. Such a strategy was intended as an alternative to IS6110 restriction fragment length polymorphism typing, which is still regarded as the reference standard for M. tuberculosis typing [13,14]. In the present study, spoligotyping was combined with IS6110-Mtb1/Mtb2 PCR [2], which has a discriminatory power close to that of IS6110 restriction fragment length polymorphism typing [15]. Two previous studies have also used IS6110-Mtb1/Mtb2 PCR for epidemiological studies of TB [15,16]. Overall, based on identical spoligotyping and IS6110-Mtb1/Mtb2 banding patterns, 12 M. tuberculosis isolates were assigned to four clonally related groups, suggesting that 25% of the TB cases studied resulted from recent transmission events. However, clustering of M. tuberculosis isolates does not always represent recent transmission, and can also reflect the persistence of well-conserved endemic strains [17]. The absence of epidemiological data to confirm a clonal relationship among the isolates was an important limitation of this study.

Phenotypic characterisation revealed that mono-resistance to either isoniazid or streptomycin was most frequent, which is broadly consistent with other findings worldwide [18]. Disturbingly, eight of the isolates, including the Beijing isolates, were MDR. However, in most cases, multidrug resistance status was probably a result of incomplete or inadequate treatment rather than an ongoing transmission of MDR strains. Indeed, none of the MDR isolates formed part of definite clonally related clusters. Two of the clusters were composed exclusively of isoniazid-resistant isolates; one contained only streptomycin-resistant isolates and one contained two isoniazid-resistant isolates and one rifampicin-resistant isolate. These observations favour a close relationship among M. tuberculosis isolates belonging to the same cluster, suggesting transmission of the same drug-resistant strain among the patients within the cluster [5].

In conclusion, a combination of two PCR-based methods, spoligotyping and IS6110-Mtb1/Mtb2, was shown to be useful for determining the relatedness among M. tuberculosis isolates and defining epidemiological links in central Poland. Nevertheless, to better determine the utility of this approach for studying the molecular epidemiology of TB, a larger pool of isolates from geographically distinct regions should be analysed.


The results included in this study were presented, in part, at the 17th European Congress of Clinical Microbiology and Infectious Diseases (Munich, 2007). Some of the data have been published previously (in Polish) in Pneumonol Alergol Pol 2007; 75: 22–31. The authors would like to thank S. Lock for kindly reviewing the manuscript. No information has been provided by the authors concerning the existence or absence of conflicting or dual interests.