Original Report

DNA Fingerprinting as an Indicator of Active Transmission of Multidrug-Resistant Tuberculosis in Poland Anna Sajduda, PhD;*, Jaroslav Dziadek, PhD;+ Agnieszka Dela, MSc;* Natalia Zalewska-Schanthaler, PhD;* Zofia Zwolska, PhD;* and Johnjoe McFadden, PhDs

ABSTRACT

Objectives: To use genetic fingerprinting to investigate the epi- demiology of tuberculosis (TB) caused by Mycobacterium fuber- culosis in Poland, a country with a relatively high incidence of tuberculosis, to improve TB control.

Design: One hundred M, tuberculosis isolates from 98 patients in the Institute of Tuberculosis and Lung Diseases in Warsaw from 1993 to 1995 and 85 isolates obtained from 50 patients in the Hospital of Lung Diseases in Lodz in 1996 were sub- jected to DNA restriction fragment length polymorphism (RFLP) analysis, using the insertion sequence IS61 IO as a probe.

Results: IS61 lo-associated banding patterns of the M. tuber- culosis isolates originating from different localities varied con- siderably, but isolates from Lodz had a higher degree of similarity. Strains with identical RFLP types were identified in patients of the same family or patients living in the same area, indicating active transmission. Of strains isolated in Warsaw, 45% were resistant to at least one drug, and 35% were resis- tant to two or more drugs and were classified as multidrug- resistant (MDR). Some drug-resistant isolates were found to have identical banding patterns and originated from epidemi- ologically linked cases.

Conclusions: Active transmission of TB, including MDR TB, is occurring in Poland. Active measures must be taken to prevent the spread of drug-resistant TB in Poland and potentially, the rest of Europe.

*Department of Genetics of Microrganisms, University of Lodz, Lodz, Poland; 7Center of Microbiology and Virology, Polish Academy of Sci- ences, Lodz, Poland; *Institute of Tuberculosis and Lung Diseases, War- saw, Poland; and “School of Biological Sciences, University of Surrey, Guildford, United Kingdom.

The work performed in Poland was supported by a grant from the State Committee for Scientific Research (ICON, Contract No. 4P05D05412). Dr. Sajduda was supported by a British Council Fellowship at the Univer- sity of Surrey, United Kingdom.

Received: February 10, 1998; Accepted: April 3, 1998.

Address correspondence to Dr. Johnjoe McFadden, School of Biological Sciences, University of Surrey, Guildford, Surrey GU2 5XH, UK.

Keywords: drug resistance, epidemiology, IS61 10, Mycobacterium tuberculosis, RFLP typing

Int J Infect Dis 1998; 3:12-17.

In April 1993, the World Health Organization declared tuberculosis (TB) “a global health emergency.“l Tubercu- losis is back as an important threat to public health, owing to a complexity of factors, such as poverty, acquired immunodeficiency syndrome (AIDS), drug addic- tion, homelessness, and changes in public health policy. The emergence of multidrug resistant (MDR) strains of the causative agent Mycobacterium tuberculosis poses an additional threat of potentially returning TB control to the ineffective treatments of the pre-antibiotic era.

The analysis of tuberculous transmission and the tracing of the sources of infection require the ability to discriminate among M. tuberculosis strains. Several methods have been used for typing these strains. Until recently, all the methods relied on the study of pheno- typic traits, such as serotype, biotype, bacteriophage sen- sitivity, or sensitivity to antimicrobial agents. These techniques have intrinsic limitations. For example, phe- notypic expression may vary with culture conditions, and there is only a small number of phage types. To over- come these difficulties, genetic markers for the identifi- cation of strains within the M. tuberculosis complex have been described. Of the five repetitive elements that so far have been discovered in M. tuberculosis complex strains, the insertion element IS6110 has proved to be the most useful marker for epidemiologic studies.z-4 Restriction fragment length polymorphism (RFLP) analy- sis based on IS61 IO has been useful for identifying trans- mission during outbreaks, 5-7 for distinguishing between new infection and reactivation in patients who relapsed after treatment,s and for the demonstration of cross-con- tamination of cultures in laboratories9 Furthermore, an internationally accepted consensus for the standardized use of IS61 10 DNA fingerprinting of M. tuberculosis strains has been established.‘O

In this study, M. tuberculosis isolates from Polish TB patients between 1993 and 1996, were typed by the RFLP method, using a probe for IS6110, and drug-resistance profiles also were obtained. Evidence was found of active

12

DNA Fingerprinting and Drug-Resistant TB in Poland / Sajduda et al 13

transmission of TB in Poland, including TB caused by MDR strains of M. tuberculosis.

MATERIAL AND METHODS

Bacterial Isolates and Clinical Data

One hundred isolates of M. tuberculosis analyzed in this study were collected at the Department of Microbiology at the Institute of Tuberculosis and Lung Diseases in War- saw, Poland. This laboratory serves as a diagnostic ser- vice laboratory for Central Poland and as a reference laboratory for all of the country. The isolates were ran- domly sampled from late 1993 to the end of 1995. The remaining 85 isolates were obtained in 1996 in the Hos- pital of Lung Diseases in Lodz. Data on the patients con- cerning type of specimen, date of isolatio.n, clinical diagnosis, TB status (pulmonary versus extrapulmonary), human immunodeficiency virus (HIV) status, chemother- apeutic regimens, and standard demographic data were registered.

The 185 isolates were isolated from sputum (165 specimens), bronchoalveolar fluid (n = 4), bronchoalve- olar lavage (n = 2), bronchial aspirate (n = 3), pleural fluid (n = 2), pleural lavage (n = l), stomach lavage (n = 4), urine (n = 2), throat swab (n = l), and wound swab (n = 1). However, in some cases in whiclh multiple isolates were from the same patient, each of the strains was identical by RJYLP typing to the first isolate. There- fore, only 148 isolates were included (one for each patient) in this study.

The species identification of the isolates Ywas based on standard microbiologic tests. All the isolates were examined for their susceptibility to isoniazid, strepto- mycin, ethambutol, and rifampin. The susceptibilities of the 100 isolates from Warsaw to para-aminosalicylic acid (PAS), ofloxacin, cycloserine, ethionamide, and capre- omycin also were determined, as described previously.ll

DNA Fingerprinting

Total genomic DNA from each M. tuberculosis isolate was prepared essentially as described elsewhere,12 and digested with PvuII for ISGIIO-RFLP and IS18081-RFLP, and with AZuI for direct repeat (DR)-RFLP Southern blot experiments were performed as described previ- ously.* Three DNA probes were used: (1) IS6110 probe, a 245-bp, polymerase chain reaction (PCR)-amplified DNA fragment generated from the 5’ end of IS61Z012; (2) IS3081 probe, a 236-bp product amplified by PCR12; and (3) a DR-r probe, a 36-bp oligonucleotide flanking the IS6110 element.13 The DNA probes were nonradio- actively labelled with a DIG DNA labelling kit (Boehringer Mannheim, Mannheim, Germany). The

PvuII-digested chromosomal DNA of M. tuberculosis Mt14323 was used as a reference in each analysis.”

Analyses of Restriction Fragment Length Polymorphism Patterns

The DNA fingerprint patterns of the isolates were com- pared both by computer-assisted analyses, using the Win- dows versions of Gelcompar (version 3.lb, Applied Maths, Kortrijk, Belgium) and ScanPack (version 3.0, Bio- metra, Germany), as well as by visual examination. Com- parisons of patterns were done by the unweighted pair group method, using arithmetic averages (unweighted- pair group analysis [UPGMA] clustering method), by using the Dice coefficient according to the instructions of the manufacturers (Applied Maths). The clustering generated was consistent with visual inspection of the data. In the few cases where multiple isolates were obtained from the same patient, they each were found to give identical RFLP patterns.

RESULTS

Patient Characteristics

One hundred eighty-five isolates from 148 patients were analyzed. Of these, 98 patients represent 24% of the TB patients examined with bacteriologic confirmation and 1% of all examined patients (positive and negative cul- tures) in the Institute of Tuberculosis and Lung Diseases in Warsaw between 1993 and 1995. The remaining 50 patients represent 21% of the patients notified with bac- teriologic confirmation of TB and 11% of all investigated patients in the Hospital of Lung Diseases in Lodz in 1996 (Table 1).

Most of the patients were of Polish origin (98.6%); only two patients (1.4%) were immigrants from foreign countries: one from Ukraine and one from Armenia. The patient ages were between 17 and 90 years; about 71% of the patients were between 20 and 60 years of age. Most of the patients (71.4%) were men. None were HIV positive.

Table 1. Numbers of Patients with TB Notified in Poland in 1993-I 996 and Enrolled in this Study

Number of Bacteriologic RFLP Analysis Patients Confirmation of M. tuberculosis

Year Investigated of TB Strains

Institute of Tuberculosis and Lung Diseases, Warsaw 1993 3079 123 7 1994 2919 126 30 1995 2887 158 61 Total 8885 407 98

Hospital of Lung Diseases, Lodz 1996 460 242 50

14 International Journal of Infectious Diseases / Volume 3, Number 1, July-September 1998

Table 2. Drug Resistance Profiles of M. tuberculosis Strains from the Institute of Tuberculosis and Lung Diseases in Warsaw*

Number of Resistant Strains Isolated

Group+ Drug 7994 (%) 1995 (%) Total (‘96)

I SM 0 4 (6.6) 4 (4.4) INH 0 4 (6.6) 4 (4.4) RMP 0 1 (1.6) 1 (1.1)

II INH, RMP 1 (3.3) 2 (3.3) 3 (3.3) SM, RMP 1 (3.3) 1 (1.6) 2 (2.2) SM, ETA 0 1 (1 .6) 1 (1.1)

Ill PAS, SM, INH 1 (3.3) 1 (1.6) 2 (2.2) PAS, SM, OFL 0 1 (1.6) 1 (1.1) RMP, SM, INH 0 7 (11.5) 7 (7.7) ETA, SM, INH 0 1 (1.6) 1 (1.1) RMP, EMB, INH 0 2 (3.3) 2 (2.2)

IV Resistant to more than 3 drugs 1 (3.3) 12 (19.7) 13 (14.3) Total 4/30 (13.3) 37/61 (60.7) 41191 (45.1)

*A total of 91 strains were tested (30 from 1994 and 61 from 1995). Drug resistance patterns of seven strains from 1993 were not available. +Groups I, II, III, and IV include strains resistant to 1, 2, 3, or more than 3 drugs, respectively. SM = streptomycin; INH = isoniazid; ETA = ethionamide; RMP = rifampin; EMB = ethambutol; PAS = para-aminosalicylic acid; OFL = ofloxacin.

Strain Characteristics 1~6110 Copy Numbers

Drug Resistance Patterns

The drug resistance patterns of 91 isolates obtained in the Institute of Tuberculosis in Warsaw are given in Table 2. Forty-one (45.1%) isolates were found to be resistant to at least one antituberculous drug, and 32 isolates (35.2%) were resistant to two or more drugs. There was a large increase in the number of resistant isolates identified in 1995 (60.7%) compared to 1994 (13.3%) and most of these (75%) were MDR strains. Among 50 isolates iso- lated in 1996 in the Hospital of Lung Diseases in Lodz only one isolate (2%) was observed to be resistant to isoniazid.

Kb I 2 3 4 5 6 7 -8 9 10 41 42r1314 45 ‘16 17 AA

Figure 1. DNA fingerprints of 16 M. tuberculosis strains originating from Poland. Southern blots of Pvull-digested chromosomal DNA hybridized with 245bp PCR fragments of IS61 IO. Numbers on the left indicate the sizes of standard DNA fragments (in kilobase pairs). Lane 18 represents DNA of M. tuberculosis reference strain Mt14323.

The copy number of IS6110 in each of the isolates was determined from the number of bands hybridizing to the probe, and ranged from 0 to 18. Figure 1 shows IS6110 RFLP patterns of 16 clinical isolates of M. tuberculosis. The DNA extracted from two isolates obtained from Pol- ish patients did not hybridize with IS61 10 DNA. Also, by PCR with the INS1 and INS2 primers5 no amplifiable IS6110 DNA was found in these strains. Hybridization patterns consistent with M. tuberculosis with the IS1081 and DR-r probes were demonstrated for both of these strains. The majority, 126 of the 148 investigated isolates (85%) contained between 6 and 11 copies of ISGZIO. None of the strains carried only a single copy of the element.

DNA Polymorphism

As shown in Table 3, 91 different DNA fingerprint pat- terns were observed in the 96 isolates from Warsaw (the two isolates carrying no IS6110 element were excluded). Of the 91 patterns, only four were shared by two or three isolates, whereas the remaining 87 isolates (90.6% of iso- lates) gave unique fingerprint patterns. The relatively low

Table 3. Polymorphisms of M. tuberculosis Isolates from 146 Patients with Tuberculosis in Poland (1993-1996)*

Unique Number of Number Year Strains Clusters in Clusters

Institute of Tuberculosis and Lung Diseases, Warsaw 1993 5 1 2 1994 27 1 2 1995 55 2 2-3 Total 87 4 5-6

Hospital of Lung Diseases, Lodz 1996 28 9-10 2-4

*Excluding two isolates carrying no IS61 IO element.

Total

7 29 60 96

50

DNA Fingerprinting and Drug-Resistant TB in Poland / Sajduda et al 15

level of clustering (9.4%) likely reflects the wide geo- graphic area from which these strains were derived. In contrast, among the 50 strains isolated in Lodz in 1996, only 38 different DNA fingerprint patterns we:re revealed (see Table 3). Of these 38 RFLP types, nine were shared by two strains and one was identical for four strains. The remaining 28 patterns (56% of strains) were unique. The level of clustering of 44% was high compared to that of the Warsaw strains, indicating that active transmission is occurring in defmed localities within Poland, such as the city of Lodz. To further investigate those strains’ with iden- tical RFLP patterns, place of residence of patients was examined. Of the four clusters of strains with identical RFLP patterns in the Warsaw strain collection, two clus- ters were found to have originated from single localities. The most prevalent pattern involved three strains iso- lated in 1995 from patients originating from a single town in Silesia (southern Poland). Two of the three isolates were obtained from members of the same family (one patient and his wife), indicating that active transmission of the infection resulted from family contact. These two isolates were also identical in the analysis of the poly- morphism in the region containing the DR elements, whereas the third member of this cluster exhibited a dif- ferent DR RFLP pattern. This observation is in agreement with previously published data showing that M. tuber- culosis strains that share the same IS6210 RFLP type can differ in their DR RFLP patterns.13 Of the three pairs of strains with identical IS6110 RFLP types, one (isolated in 1993) was identified in two patients originating from a single town in Mazovia (eastern Poland). These strains were also identical by DR fingerprinting. The other two pairs of identical strains were from patients with no apparent epidemiologic links. A higher numbler of clus- tered strains (22 versus 9) were found among the iso- lates obtained in a local hospital in Lodz. In a few cases the strains with identical DNA fingerprint patterns were isolated from patients living in the same district of the town, indicating active transmission of the disease. None of the patients belonging to these chains of transmission were of foreign origin, and there was no significant age difference between patients with unique strains and those with clustered strains.

Investigation of Correlation between ResB-iction Fragment Length Polymorphism Types and Drug Resistance

IS6110 DNA fmgerprints of 42 drug-resistant isolates were compared with those of 99 drug-sensitive isolates. The results showed no noticeable difference in banding pat- tern or copy number of 1~6110 between drug-sensitive and drug-resistant isolates. Susceptible and resistant or even MDR strains were found in the same clusters. One clone of three MDR isolates was also identified. Two of the isolates exhibited identical drug resistance profiles and came from the epidemiologically linked TB cases

from Silesia. The third isolate of this clone was isolated from a patient without apparent epidemiologic relations with the two other patients, was resistant to two extra drugs (ethambutol and capreomycin), and showed a different DR-RFLP pattern.

DISCUSSION

Poland is a central European country with 38.6 million inhabitants (1997). The social security system is mainly public and is financed through taxes. A monitoring and prevention policy for TB was implemented in 1957. The general standard of living has improved during the post- war decades, up until the 1980s when a significant dete- rioration of socioeconomic conditions was observed due to substantial political and economic changes. The patient notification rate was 39.8 per 100,000 in 1996.‘* For com- parison, the incidence of TB in different regions in France ranged between 9 and 36 per 100,000 in 1991,15 and was as low as 5.4 per 100,000 in Denmark in 1992.‘”

IS6110 DNA fingerprinting demonstrated that the degree of DNA polymorphism of Polish M. tuberculosis strains examined in this study was high. The average num- ber of isolates per pattern was 1.05 for the strains from Warsaw, whereas for the recent study in NewYork it was 1.35,17 and for the study in San Francisco, it was 1.5.18 It might be expected that regions suffering from a high level of TB transmission and insufficient hospital facilities, with a high risk of nosocomial infections, would be char- acterized by a relatively limited variety of mycobacterial clones. The diversity of patterns observed in the present study may be explained by the fact that the majority of analyzed strains were isolated from TB patients originat- ing from different regions of the country. On the other hand, the 50 strains isolated in the Hospital of Lung Dis- eases in Lodz showed a relatively high level of clustering (44%) and a corresponding low level of DNA polymor- phism (number of isolates per pattern was 1.32). The lev- els of clustering are similar to those described recently in New York and San Francisco.“zls Lodz is the second biggest city in Poland, with about 839,000 inhabitants (in 1997). It is an important industrial, trade, and communi- cation center. Despite the mobility of the local population during the last few decades, the authors found a signifi- cant level of clustering of strains, indicating active trans- mission of TB in this area. The results indicate that, in contrast to most western European countries, active tmns- mission among HIV-negative patients is a significant fac- tor in the epidemiology of TB in Poland and indicates the need for better control measures to prevent spread of infection.

In accordance with recent observations from a large- scale RFLP study of M. tuberculosis isolates in Tunisia,1o French Polynesia,‘* and Denmark,16 the majority of iso- lates investigated in the present study (85%) carried from 6 to 11 copies of the IS6110 element. Two isolates were

16 International Journal of Infectious Diseases / Volume 3, Number I, July-September 1998

found not to carry any detectable IS6110 element. Hybridization experiments with other genetic markers revealed the presence of DR and IS1081 DNAs in the genomes of these strains. This observation is unique for European populations, as to date such cases have been reported only for M. tuberculosis strains isolated from patients originating from Asia, mainly from India and Viet- nam.3s19 The absence of IS6110 in Polish isolates of M. tuberculosis indicates that diagnosis of TB based on PCR- amplification of IS6110 may lead to false-negative results for some European isolates.5

Resistance to anti-TB drugs can develop as a result of treatment (secondary resistance), but it also can occur as a result of infection with drug-resistant strains of M. tuberculosis (primary resistance). In the present study, drug resistance to one or more anti-TB drugs was found in 45% of the isolates obtained in the Institute of Tuber- culosis and Lung Diseases in Warsaw. The MDR isolates (resistant to two or more drugs) accounted for 35.2% of the total number of those isolates but 75% of isolates from 1995. The majority of the isolates analyzed in the present study were isolated in the Institute of Tubercu- losis and Lung Diseases in Warsaw where selected TB cases are treated (e.g., those difficult to diagnose, uncared- for in treatment, or patients with chronic tuberculous disease).” Thus, the results are not representative for the whole Polish population. According to the first large-scale study of drug-resistant TB in Poland, 6.9% of M. tubercu- losis isolates were resistant (Professor Zofia Zwolska, World Health Organization/International Union against Tuberculosis and Lung Disease [IUATLD] global project on anti-tuberculosis drug resistance surveillance. Poland 1996- 1997. Unpublished results). Monoresistance to iso- niazid and streptomycin was predominant (2.4% and 1.6%, respectively). The isolates resistant to two drugs accounted for 1.3% (resistance to streptomycin and iso- niazid was predominant in this group and accounted for 0.8%) whereas 1.5% of all the tested isolates were resis- tant to three anti-TB drugs. In this latter group resistance to streptomycin, isoniazid, and rifampin was predominant (1.2%) and 0.4% of the isolates were resistant to all four drugs tested (streptomycin, isoniazid, rifampin, and etham- butol). The increasing frequency of MDR strains from 1994 to I995 in the present study may reflect regional differences in the rate of MDR TB in Poland but, never- theless, is cause for concern, and indicates that the situ- ation should be carefully monitored.

In accordance with previous findings,15sz0 no gross differences in RFLP patterns were observed between drug-sensitive and drug-resistant isolates. However, a few clones of drug-resistant strains were found to infect small groups of epidemiologically linked patients. This indi- cates that active transmission of drug-resistant M. tuber- culosis strains is taking place in Poland. This observation is of considerable importance in terms of geographic

localization of Poland in the center of Europe, and increas- ing migration across country borders that facilitates the transfer of infectious, airborne diseases like tuberculosis.

In conclusion, the typing of isolates isolated from TB patients in Poland over 3 years provided insight into the status of tuberculosis in this region. The results indicate that active transmission of M. tuberculosis, including drug- resistant strains, is taking place in Poland and that urgent control measures may be necessary to prevent the spread of drug-resistant TB within Poland and, potentially, the rest of Europe.

ACKNOWLEDGMENTS

The authors thank Dick van Soolingen for supplying the refer- ence strain and Jeremy W Dale for his help with the computer analysis.

REFERENCES

1. World Health Organization. TB-a global emergency. WHO report on the TB epidemic. Geneva: World Health Organi- zation, 1994.

2. van Embden JDA, van Soolingen D, Small PM, Hermans PWM. Genetic markers for the epidemiology of tuberculosis. Res Microbial 1992; 143:385-391.

3. van Soolingen D, de Haas PEW, Hermans PWM, Groenen PMA, van Embden JDA. Comparison of various repetitive DNA elements as genetic markers for strain differentiation and epidemiology of Mycobacterium tuberculosis. J Clin Microbial 1993; 31:1987-1995.

4. Yang ZH., de Haas PEW, van Soolingen D, van Embden JDA, Andersen AB. Restriction fragment length polymorphism of Mycobacterium tuberculosis strains isolated from Green- land during 1992. Evidence of tuberculosis transmission between Greenland and Denmark. J Clin Microbial 1994; 32:3018-3025.

5 Hermans PWM, van Soolingen D, Dale JW et al. Insertion element IS986 from Mycobacterium tuberculosis: a useful tool for diagnosis and epidemiology of tuberculosis. J Clin Microbial 1990; 28:2051-2058.

6. van Soolingen D, Hermans PWM, de Haas PEW; So01 DR, van Embden JDA. Occurrence and stability of insertion sequences in Mycobacterium tuberculosis complex strains: evaluation of an insertion sequence-dependent DNA poly- morphism as a tool in the epidemiology of tuberculosis. J Clin Microbial 1991; 29:2578-2586.

7. Daley CL, Small PM, Schecter GE et al. An outbreak of tuber- culosis with accelerated progression among persons infected with the human immunodeficiency virus. N Engl J Med 1992; 326:231-235.

8. Otal I, Martin C, L&y-F&bat& V, Thierry D, Gicquel B. Restric- tion fragment length polymorphism analysis using ISGllO as an epidemiological marker in tuberculosis. J Clin Micro- biol 1991; 29:1252-1254.

9. Small PM, McClenny NB, Singh SE Schoolnik GK, Tompkins LS, Mickelsen PA. Molecular strain typing of Mycobacterium tuberdosis to confirm cross-contamination in the mycobac- teriology laboratory and modification of procedures to min- imize occurrence of false-positive cultures. J Clin Microbial 1993; 31:1677-1682.

DNA Fingerprinting and Drug-Resistant TB in Poland / Sajduda et al 17

10. van Embden JDA, Cave MD, Crawford JT, et al. Strain identi- fication of Mycobacterium tuberculosis by DNA finger- printing: recommendations for the standardized methodology. J Clin Microbial 1993; 31:406-409.

11. Zalewska-Schonthaler N, Jezierska-Anczukow A. Analysis of drug sensitivity of M. tuberculosis strains isolated from patients treated because of tuberculosis in 1982 - 1991. Pneu- mono1 Pol 1993; 61:254-262.

12. van Soolingen D, de Haas PEW Hermans PWM,van Embden JDA. DNA fingerprinting of Mycobacterium tuberculosis. Methods Enzymol 1994; 235:196-205.

13. Hermans PWM, van Soolingen D, Bik EM, de Haas PEW, Dale JW, van Embden JDA. Insertion element IS987 from Mycobacterium bovis BCG is located in hot-spot integration region for insertion elements in Mycobacterium tubercu- Zosis complex strains. Infect Immun 1991; 59:2695-2705.

14. Tuberculosis and Lung Disease Institute, Department of Epi- demiology Tuberculosis and lung diseases in Poland in 1996. Report of the Institute of Tuberculosis and Lung Diseases, Warsaw, Poland, 1997.

15. Torrea G, Levee G, Grimont P Martin C, Chanteau S, Gicquel B. Chromosomal DNA fingerprinting analysis using the inser- tion sequence IS6220 and the repetitive element DR as

strain specific markers for epidemiological study of tuber- culosis in French Polynesia. J Clin Microbial 1995; 33:1899-1904.

16. Yang ZH, de Haas PEW, Wachmann CH, van Soolingen D, van Embden JDA, Andersen &i. Molecular epidemiology of tuberculosis in Denmark in 1992. J Clin Microbial 1995; 33:2077-2081.

17. Alland D, Kalkut GE, Moss AR, et al. Transmission of tuber- culosis in New York city, an analysis by DNA fingerprinting and conventional epidemiological methods. N Engl J Med 1994; 330:1710-1716.

18. Small PM, Hopewell PC, Singh SP, et al. The epidemiology of tuberculosis in San Francisco, a population-based study using conventional and molecular methods. N Engl J Med 1994; 30:1703-1709.

19. Yuen LKW Ross BC, Jackson KM, Dwyer B. Characterization of Mycobacterium tuberculosis strains from Vietnamese patients by Southern blot hybridization. J Clin Microbial 1993; 31:1615-1618.

20. Chevrel-Dellagi D, Abderrahman A, Haltiti R, Koubaji H, Gic- quel B, Dellagi K. Large-scale DNA fingerprinting of Mycobac- terium tuberculosis as a tool for epidemiological studies of tuberculosis. J Clin Microbial 1993; 31:2446-2450.