Epidemic spread of multidrug-resistant (MDR) tuberculosis in Johannesburg, South Africa

Transcription

1 JCM Accepts, published online ahead of print on 3 April 203 J. Clin. Microbiol. doi:0.28/jcm Copyright 203, American Society for Microbiology. All Rights Reserved. Epidemic spread of multidrug-resistant (MDR) tuberculosis in Johannesburg, South Africa Ben J. Marais BJ, Charmaine K. Mlambo 2, Nalin Rastogi 3, Thierry Zozio 3, Adriano G. Duse 2, Thomas C. Victor T 4, Else Marais 2, Robin M. Warren 4* Affiliations: Sydney Emerging Infectious Diseases and Biosecurity Institute (SEIB), the University of Sydney, Sydney, Australia. 2 Division of Clinical Microbiology and Infectious Diseases, School of Pathology, of the University of the Witwatersrand and National Health Laboratory Service, Johannesburg, South Africa. 3 WHO Supranational TB Reference Laboratory, TB and Mycobacteria Unit, Institut Pasteur de la Guadeloupe, Abymes, France. 4 DST/NRF Centre of Excellence for Biomedical Tuberculosis Research/MRC Centre for Molecular and Cellular Biology, Division of Molecular Biology and Human Genetics, Faculty of Medicine and Health Sciences, Stellenbosch University, Tygerberg, South Africa. *Corresponding Author: DST/NRF Centre of Excellence for Biomedical Tuberculosis Research Division of Molecular Biology and Human Genetics Faculty of Medicine and Health Sciences Stellenbosch University P.O. Box 9063, Tygerberg, 7505 South Africa address: rw@sun.ac.za Telephone: Running Title: Epidemic spread of MDR-TB Key words: Multidrug-resistant (MDR), tuberculosis, epidemic spread, transmission Word Count: 2353 Downloaded from on September 4, 208 by guest

2 29 Conflict of Interest: The authors declare no conflict of interest. Downloaded from on September 4, 208 by guest

3 ABSTRACT [Word Count: 228] Numerous reports have documented isolated transmission events or clonal outbreaks of multidrug-resistant Mycobacterium tuberculosis strains, but knowledge of their epidemic spread remains limited. In this study, we evaluated drug resistance, strain diversity and clustering rates in patients diagnosed with multidrug-resistant (MDR) tuberculosis (TB) at the National Health Laboratory Service (NHLS) Central TB laboratory in Johannesburg, South Africa, between March 2004 and December Phenotypic drug susceptibility testing was done using the indirect proportion method while each isolate was genotyped using a combination of spoligotyping and 2- MIRU typing. Isolates from 434 MDR-TB patients were evaluated, of which 238 (54.8%) were resistant to four first-line drugs (isoniazid, rifampicin, ethambutol and streptomycin). Spoligotyping identified 56 different strains and 28 clusters of variable size (2-7 cases per cluster) with a clustering rate of 87.%. Ten clusters included 337 (77.6%) of all cases, with strains of the Beijing genotype being most prevalent (6.4%). Combined analysis of spoligotyping and 2-MIRU typing increased the discriminatory power (HGDI=0.962) and reduced the clustering rate to 66.8%. Resolution of Beijing genotype strains were further enhanced when using the 24-MIRU-VNTR typing method by identifying 5 sub-clusters and 6 unique strains from twelve 2-MIRU clusters. High levels of clustering among a variety of strains suggest true epidemic spread of MDR-TB in the study setting, emphasizing the urgency of early diagnosis and effective treatment to reduce transmission within this community. Downloaded from on September 4, 208 by guest

4 INTRODUCTION The emergence of drug-resistant strains of Mycobacterium tuberculosis was documented soon after the first tuberculosis (TB) drugs became available in the late 940s (). The observation that this could be minimized with adherent, quality assured multi-drug therapy, provided strong motivation for the implementation of the DOTS strategy in the early 990s (2). The strategy aimed to achieve global TB control through microscopic diagnosis of infectious cases, standardized multidrug therapy, a reliable supply of quality assured drugs and ensuring treatment adherence through directly observed therapy (DOT). World-wide implementation of the strategy has cured some 50 million patients and averted 7 million deaths compared to pre-dots standards (3), but has been unable to stem the disconcerting rise of drug-resistant TB. Descriptions of the fitness cost associated with the acquisition of drug resistance (4), may explain why public health specialists failed to appreciated the potential for epidemic spread of drugresistant TB. Limited resources, absent laboratory infrastructure and unavailability of second-line drugs motivated the DOTS strategy s select focus on sputum smear-positive cases and treatment with standardized first-line regimens. Despite intermittent reports of patients with drug-resistant TB, the expectation was that these would disappear once program performance improved and the proverbial tap was turned off. Sustained epidemic spread of drug-resistant TB was considered unlikely (5). However, the description of a large cluster of extremely drug-resistant (XDR)-TB cases in South Africa provided stark evidence that some drug-resistant strains are highly transmissible, at least among immune compromised patients (6); similar to an earlier outbreak of multidrug-resistant (MDR)-TB in New York city (7). This prompted a prominent focus on drug-resistant TB in the revised Stop TB Strategy (8), with clear description of the critical steps required for its prevention and control (9). The global burden of drug-resistant TB remains poorly quantified. The World Health Organization (WHO) estimated that in 20 approximately (5.3%) of the 2 million prevalent TB cases had MDR-TB (0). Only a small fraction of reported cases (<20%) were correctly diagnosed and Downloaded from on September 4, 208 by guest

5 even fewer were treated according to WHO standards. This reflects major structural, economic and political constraints (). More than 80 countries have reported cases of extremely drug-resistant (XDR)-TB (0), and the drug-resistant TB disease burden in sub-saharan Africa is likely to be grossly under-estimated (2). Southern Africa is the epicenter of the dual human immunodeficiency virus (HIV)/TB epidemic, which has been associated with high rates of drug-resistant TB (0). These alarming figures emphasize the need for improved epidemiological understanding, including better description of the molecular epidemiology and transmission dynamics of MDR-TB. Molecular tools provide the means to assess the evolution and spread of M. tuberculosis. High resolution methods demonstrated the important contribution of exogenous re-infection to TB recurrence, especially in endemic areas (3). Geospatial mapping of case clusters, and whole genome sequencing combined with detailed social network analysis, have been used to guide public health responses (4). Spoligotyping provides a simple standardized method of strain typing and has enhanced our understanding of global strain distribution (5). Strains of the Beijing genotype have a very distinct spoligotype, but this monomorphic signature reduces spoligotyping s discriminatory power. Additional genetic markers such as multiple interspersed repetitive unit (MIRU)-typing provide enhanced resolution (6). Although molecular typing is routinely performed in most developed countries (7), it is rarely done in TB endemic areas where the information would be most informative. High and rising rates of drug-resistant TB have been recorded in South Africa, with some welldefined clonal outbreaks (6,8). However, strain diversity among MDR-TB cases has not been assessed in Johannesburg, the country s most populous city and cultural melting pot. We aimed to perform a detailed molecular epidemiological assessment of MDR-TB cases within the greater Johannesburg area and to evaluate evidence for epidemic spread within the study setting. Downloaded from on September 4, 208 by guest

6 METHODS Setting and study design Johannesburg s population exceeds 3 million people, with nearly 50% being under 35 years of age and growing at an estimated rate of 4 % each year (9). People migrate to Johannesburg in search of jobs and better living conditions from all over Africa; many reside in low-cost informal settlements that are plagued by overcrowding and poverty. This retrospective descriptive study included isolates from all MDR-TB patients (first MDR isolate) identified at the National Health Laboratory Service (NHLS) Central TB laboratory in Johannesburg, from March 2004 to December Sputum samples were submitted from over 00 hospitals and clinics. Limited demographic data were obtained retrospectively from laboratory records; HIV infection status was not recorded. The NHLS Central TB laboratory processes the vast majority of sputum samples in Johannesburg. Sample processing Mycobacteria Growth Indicator Tube cultures (BACTEC MGIT 960, Becton Dickenson BioSciences, Sparks, MD, USA) were obtained and a heat-killed aliquot collected for spoligo- and MIRU-typing. Sample processing was performed in a biosafety level 3 laboratory using standard procedures. Drug susceptibility testing (DST) was performed per physician request. Guidelines advised testing of all re-treatment cases and treatment failures, but adherence to these recommendations were not monitored. Resistance against all first-line drugs, excluding pyrazinamide (PZA), was assessed using the indirect proportional method in BACTEC MGIT 960 tubes; for isoniazid (INH) at 0.ug/ml, rifampicin (RMP) at.0ug/ml, ethambutol (EMB) at 5.0ug/ml and streptomycin (S) at.0ug/ml. Spoligotyping was performed according to the manufacturer s instructions using membranes and equipment provided with the spoligotyping kit (Isogen, Bioscience BV, Utrecht, The Netherlands) (20) Polymerase chain reaction (PCR) amplification was performed using an icycler Thermocycler (Bio-rad Hercules, CA, USA). This genotyping method determined the presence or absence of 43 direct and variable repeat sequences with the Direct Repeat region thereby generating spoligotype Downloaded from on September 4, 208 by guest

7 signatures which were characteristic defined strain genotypes (2). In addition to spoligotyping, all strains were typed by 2-MIRU using standard methods (6,22) and 24-MIRU was performed on all Beijing genotype strains (22). This method determined the number of repeat sequences present at each allele thereby generating a numerical profile for each strain. Results were compared with the SITVIT2 proprietary database of Institut Pasteur de la Guadeloupe, which is an updated multimarker version of previously released SpolDB4 (23) and SITVITWEB (24) databases. In this database, Spoligotype International Type (SIT) and MIRU International Type (MIT) designate spoligotypes and MIRU patterns shared by 2 or more patient isolates, respectively, whereas orphan designates patterns reported for a single isolate. Major phylogenetic clades were assigned according to signatures provided in previous versions (23,24). Statistical Analysis Cluster analysis was performed for spoligotyping alone and in combination with 2-MIRU typing. A cluster was defined as two or more strains with identical genetic patterns. The clustering rate was calculated as: (number of clustered cases - number of clusters)/total number of cases. The discriminatory value of various strain typing methods was calculated using the Hunter Gaston Discriminatory Index (HGDI) (25) where N is the total number of strains in the typing scheme, s is the total number of different patterns, and n j is the number of strains belonging to the j th pattern. This method was also used to assess the individual contribution of each of the 24-MIRU loci to discriminate between Beijing genotype strains. For Beijing genotype strains a minimum spanning tree was built with 2- and 24-MIRU profiles using BioNumerics software (Version 3 5, Applied Maths, Sint-Martens-Latem, Belgium) to reflect likely evolutionary relationships. Ethics approval was granted by the University of the Witwatersrand Human Research (Medical) Ethics Committee (Number M050628). Downloaded from on September 4, 208 by guest

8 RESULTS A total of 434 MDR-TB patients were identified and included in the study. The average age of adult patients was 35.7 years; children <5yrs of age comprised 7.6% (33/434) of cases. Most isolates (238/434; 54.8%) were resistant to all four first-line TB drugs tested, with the lowest number (94; 2.7%) being resistant to INH and RMP alone. Table summarizes the relevant demographic and DST data. Spoligotyping detected 56 different strains and 28 clusters of variable size (2-7 cases per cluster). Comparison with existing international databases identified 50 shared-types; 45 (47 cases) matched pre-existing shared-types and 5 ( cases) were newly created as they matched existing orphan types in the SITVIT2 database [SIT2996 (n=2), SIT2997 (n=6), SIT2998 (n=), SIT2999 (n=) and SIT296 (n=)]. Beijing genotype strains were most common, although they represented only 6.4% (7/434) of all cases. Six novel strains exhibited unique spoligotyping signatures ( each from the LAM, Haarlem (H) and S families and 3 of unknown origin). Ten spoligotype clusters included 77.6% (337/434) of all isolates. Additional 2-MIRU analysis identified 39 subclusters (2-64 isolates per cluster) and 25 non-clustering strains (Table 2). Six 2-MIRU clusters (MIT7=7; MIT246=64; MIT8=9; MIT2=35; MIT20=8; MIT224=7) included more than 0 cases. A combination of 2-MIRU and spoligotyping improved the discriminatory power (HGDI = 0.962), compared to spoligotyping alone (HGDI=0.97). The resolution increase achieved with 2- MIRU was greatest for the Beijing and Latin-America/Mediterranean (LAM) genotypes. A combination of spoligo and 2-MIRU typing yielded a clustering rate of 66.8% by spoligotyping, 87.% with spoligotyping and 2-MIRU-typing and 72.% with 2-MIRU-typing alone. Beijing genotype strains were further differentiated using 24-MIRU, which facilitated a more detailed phylogenetic reconstruction relative to the 2-MIRU method (Figure ). It also increased the resolution from 9 2-MIRU profiles (64 clustered and 7 unique) to MIRU profiles (48 clustered and 23 unique) (Table 3). Most unique strains were very similar to clustered strains, with only minor ( allele) differences. Overall, 24-MIRU provided improved discriminatory power (HGDI Downloaded from on September 4, 208 by guest

9 ) compared to 2-MIRU (HGDI 0.90). Table 4 reflects the discriminatory value contributed by each of the 24 loci. 2-MIRU included 5 loci with high discriminatory value; the MIRU3 added the greatest value (HGDI 0.6). A previously proposed 5 loci combination (22) included 7 loci with moderate to high discriminatory value while 24-MIRU included 0 loci with moderate to high discriminatory value. Loci MIRU2, MIRU4, MIRU20, MIRU23, MIRU24, and ETR-B demonstrated no variability. Downloaded from on September 4, 208 by guest

10 DISCUSSION The clustering rate of 66.8% using a combination of spoligo- and 2-MIRU typing, which has similar discriminatory power as classic IS60 restriction fragment length polymorphisms (RFLP) fingerprinting, indicates frequent MDR-TB transmission in the greater Johannesburg area of South Africa. The large diversity of MDR-TB strains and the multiple clusters identified indicate that this is not an isolated clonal outbreak. It reflects transmission from numerous source cases, confirming true epidemic spread of MDR-TB within the study community. These findings support modelling studies that predicted the likely emergence of highly transmissible MDR-TB strains over time (26). Given the slow and highly variable rate at which M. tuberculosis infection generates secondary cases (27), it is likely that transmission events were underestimated in the 3-year snap shot provided by this study. Selection and amplification of drug resistance are facilitated by exposure to poor quality drugs, sub-optimal regimens and poor treatment adherence, but may occur even in well-functioning TB control programs (26). Studies on clinical MDR-TB strains have demonstrated that the fitness cost associated with acquired drug resistance can be overcome by various compensatory mechanisms (28). The opportunity for compensatory evolution is enhanced by ongoing selective drug pressure, especially with poorly targeted therapy. This allows the best adapted drug-resistant strains to regain (or even exceed) full fitness (28). Use of a standardized re-treatment regimen remains common practice in many TB endemic areas, despite the fact that the addition of a single agent (streptomycin) in patients with TB treatment failure or relapse encourages multiplication of resistance (29). Failure to perform DST leads to poorly targeted treatment and bad outcomes despite adequate treatment adherence (29); with ongoing selection of the fittest drug resistant strains. The fact that the majority of clustered MDR strains were resistant to multiple first-line drugs indicates primary transmission of these strains. Early diagnosis and rapid initiation of effective treatment is essential to control the TB epidemic, especially in communities with high infection pressure (30). Downloaded from on September 4, 208 by guest

11 The diagnosis of young children with culture confirmed MDR-TB provides further support of ongoing transmission within the study community (3). Children with MDR-TB are likely to have been under-represented given the need for admission to a referral hospital for adequate diagnostic work-up, which limits diagnostic access in most resource-limited settings (3). High rates of MDR- TB has been documented in children from Johannesburg recruited into a prospective INH preventive therapy trial in HIV-exposed infants; confirming transmission within HIV-affected households (32). It has been postulated that drug-resistant strains with reduced fitness may be more readily transmitted in settings with a high HIV prevalence (33). Rising numbers of children diagnosed with drug-resistant TB in Cape Town, suggests increasing transmission rates in other parts of South Africa as well (34). No major gender-related differences were observed, but the young average age of adult TB patients (especially women; 33.8 years) is consistent with pronounced shifts in the age and gender profile induced by heterosexual spread of HIV infection (35). The strain diversity observed is broadly consistent with studies from other parts of South Africa (36,37), probably reflecting extensive population mixing. Interestingly, LAM-ZWE (SIT59) that predominates in neighboring Zambia and Zimbabwe was rare (38), with common strains in other parts of Africa being totally absent. This is surprising given the huge number of African immigrants that are resident in townships surrounding Johannesburg. The dominance of locally circulating strains may indicate that imported strains struggle to replace resident strains, but may also reflect some selection bias since immigrants are required to present a valid visa to access expensive second-line TB treatment services. The high prevalence of SIT60 (LAM4/F5/KZN), the strain responsible for the XDR-TB outbreak in KwaZulu Natal Province (6), reflects long standing labour migration patterns between KwaZulu Natal and gold mines in Johannesburg. This strain has already demonstrated capacity to acquire additional second-line resistance (and retain fitness to cause outbreaks), which provides a stark warning to the public health community (39). Downloaded from on September 4, 208 by guest

12 The Beijing genotype, which has a well-documented association with transmitted drug resistance (40-43), were most prevalent. Its ability to replace resident strains has been demonstrated in a community-based study in Cape Town that prospectively tracked the rapid emergence of Beijing family strains over a 2 year period (44). The success and global spread of the Beijing genotype has been ascribed to multiple factors (45), including resistance to Mycobacterium bovis BCGinduced immunity (46), potential hyper-virulence and reduced fitness costs associated with the acquisition of drug resistance (47). Resolution of Beijing genotype strains was enhanced by 24- MIRU typing relative to the 2-MIRU typing. This facilitated detailed mapping of likely transmission clusters (Figure ) while the identification of minor variants suggests that that these strains have been in circulation for an extended period and have become endemic in the region. Pronounced differences in the size of transmission clusters suggest highly variable strain-related transmissibility, even among highly monomorphic Beijing genotype strains. Improved understanding of the factors that determine the transmissibility of particular strains has high public health importance. The study is limited by selective sampling, since only specimens for which the physician requested DST were eligible for inclusion. Differential testing could have biased clustering estimates, especially if physicians were more likely to test contacts of known MDR cases or if specific facilities where transmitted (primary) MDR-TB was more of a concern were more likely to request DST. In addition, re-treatment and HIV infection status, total case numbers processed by the laboratory and strain diversity among non-mdr strains were not available in this retrospective analysis, which would have provided important epidemiological context. However, although cluster rates may have been influenced by biased sampling, the observation that MDR-TB is regularly transmitted within the study community remains valid and has key public health importance. In summary, this analysis of MDR strains identified in routine clinical practice provides convincing evidence of true epidemic spread of MDR-TB in Johannesburg, South Africa. It indicates that simply turning off the tap through improved programmatic management of drug sensitive TB is Downloaded from on September 4, 208 by guest

13 insufficient to contain the spread of drug-resistant TB, although it remains important. New strategies should include early accurate diagnosis and effective treatment of all MDR-TB cases, as well as consideration of creative public health approaches to reduce TB transmission within identified transmission hotspots (48,49). Downloaded from on September 4, 208 by guest

14 283 ACKNOWLEDGEMENTS The authors would like to thank the staff of the Tuberculosis Referral Laboratory, National Health Laboratory Services, Johannesburg, the Pasteur Institute in Guadeloupe, University of Stellenbosch (Tygerberg) and University of the Witwatersrand (CMID) for their assistance. This work was supported by the South African Tuberculosis and AIDS Training (SATBAT) program (National Institute of Health/Fogarty International Centre U2RTW007370/3), the Third World Women in Science (TWOWS), University of the Witwatersrand Health Sciences, European Regional Development Fund, European Commission (ERDF/FEDER, A34-05), and the Regional Council of Guadeloupe (Biodiversity project, CR08/03380), Medical Research Council of South Africa and the NRF/DST Centre of Excellence for Biomedical Tuberculosis Research. Dr Zozio was awarded a Ph.D. fellowship by the European Social Funds through the Regional Council of Guadeloupe. Downloaded from on September 4, 208 by guest

15 REFERENCES: Mitchison, D. and G. Davies The chemotherapy of tuberculosis: past, present and future. Int.J.Tuberc.Lung Dis. 6: Raviglione, M. C. and A. Pio Evolution of WHO policies for tuberculosis control, Lancet 359: Raviglione, M., B. Marais, K. Floyd, K. Lonnroth, H. Getahun, G. B. Migliori, A. D. Harries, P. Nunn, C. Lienhardt, S. Graham, J. Chakaya, K. Weyer, S. Cole, S. H. Kaufmann, and A. Zumla Scaling up interventions to achieve global tuberculosis control: progress and new developments. Lancet 379: Middlebrook, G. and M. L. COHN Some observations on the pathogenicity of isoniazid-resistant variants of tubercle bacilli. Science 8: Dye, C. and M. A. Espinal Will tuberculosis become resistant to all antibiotics? Proc.R.Soc.Lond B Biol.Sci. 268: Gandhi, N. R., A. Moll, A. W. Sturm, R. Pawinski, T. Govender, U. Lalloo, K. Zeller, J. Andrews, and G. Friedland Extensively Drug Resistant Tuberculosis as a cause of death in patients co-infected with Tuberculosis and HIV in a rural area of South Africa. Lancet 368: Moss, A. R., D. Alland, E. Telzak, D. Hewlett, Jr., V. Sharp, P. Chiliade, V. LaBombardi, D. Kabus, B. Hanna, L. Palumbo, K. Brudney, A. Weltman, K. Stoeckle, K. Chirgwin, M. Simberkoff, S. Moghazeh, W. Eisner, M. Lutfey, and B. Kreiswirth A city-wide outbreak of a multiple-drug-resistant strain of Mycobacterium tuberculosis in New York. Int.J.Tuberc.Lung Dis. :5-2. Downloaded from on September 4, 208 by guest

16 WHO Stop TB Strategy. Available at: Accessed Nathanson, E., P. Nunn, M. Uplekar, K. Floyd, E. Jaramillo, K. Lonnroth, D. Weil, and M. Raviglione MDR tuberculosis--critical steps for prevention and control. N.Engl.J.Med. 363: WHO. 202Global Tuberculosis Report 202. Available at: Accessed Zumla, A., I. Abubakar, M. Raviglione, M. Hoelscher, L. Ditiu, T. D. McHugh, S. B. Squire, H. Cox, N. Ford, R. McNerney, B. Marais, M. Grobusch, S. D. Lawn, G. B. Migliori, P. Mwaba, J. O'Grady, M. Pletschette, A. Ramsay, J. Chakaya, M. Schito, S. Swaminathan, Z. Memish, M. Maeurer, and R. Atun Drug-resistant tuberculosis-- current dilemmas, unanswered questions, challenges, and priority needs. J.Infect.Dis. 205 Suppl 2:S228-S Ben, A. Y., B. Nemser, A. Singh, A. Sankin, and N. Schluger Underreported threat of multidrug-resistant tuberculosis in Africa. Emerg.Infect.Dis. 4: Verver, S., R. M. Warren, N. Beyers, M. Richardson, G. D. van der Spuy, M. W. Borgdorff, D. A. Enarson, M. A. Behr, and P. D. van Helden Rate of reinfection tuberculosis after successful treatment is higher than rate of new tuberculosis. Am.J.Respir.Crit Care Med. 7: Gardy, J. L., J. C. Johnston, S. J. Ho Sui, V. J. Cook, L. Shah, E. Brodkin, S. Rempel, R. Moore, Y. Zhao, R. Holt, R. Varhol, I. Birol, M. Lem, M. K. Sharma, K. Elwood, S. J. Jones, F. S. Brinkman, R. C. Brunham, and P. Tang. 20. Whole-genome sequencing and social-network analysis of a tuberculosis outbreak. N.Engl.J.Med. 364: Downloaded from on September 4, 208 by guest

17 Sola, C., I. Filliol, M. C. Gutierrez, I. Mokrousov, V. Vincent, and N. Rastogi Spoligotype database of Mycobacterium tuberculosis: biogeographic distribution of shared types and epidemiologic and phylogenetic perspectives. Emerg.Infect.Dis. 7: Supply, P., S. Lesjean, E. Savine, K. Kremer, D. van Soolingen, and C. Locht Automated high-throughput genotyping for study of global epidemiology of Mycobacterium tuberculosis based on mycobacterial interspersed repetitive units. J.Clin.Microbiol. 39: Grant, J., S. Kammerer, B. Baker, and L. Kim Tuberculosis genotyping - United States, MMWR 6: Johnson, R., R. M. Warren, G. D. van der Spuy, N. C. Gey Van Pittius, D. Theron, E. M. Streicher, M. Bosman, G. J. Coetzee, P. D. van Helden, and T. C. Victor Drugresistant tuberculosis epidemic in the Western Cape driven by a virulent Beijing genotype strain. Int.J.Tuberc.Lung Dis. 4: Statistics South Africa Mid-year population estimates Available at: Accessed Kamerbeek, J., L. Schouls, A. Kolk, M. van Agterveld, D. van Soolingen, S. Kuijper, A. Bunschoten, H. Molhuizen, R. Shaw, M. Goyal, and J. Van Embden Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology. J.Clin.Microbiol. 35: Streicher, E. M., T. C. Victor, S. G. van der, C. Sola, N. Rastogi, P. D. van Helden, and R. M. Warren Spoligotype signatures in the Mycobacterium tuberculosis complex. J.Clin.Microbiol. 45: Downloaded from on September 4, 208 by guest Supply, P., C. Allix, S. Lesjean, M. Cardoso-Oelemann, S. Rusch-Gerdes, E. Willery, E. Savine, H. P. de, D. H. van, S. Roring, P. Bifani, N. Kurepina, B. Kreiswirth, C. Sola, N.

18 Rastogi, V. Vatin, M. C. Gutierrez, M. Fauville, S. Niemann, R. Skuce, K. Kremer, C. Locht, and S. D. Van Proposal for standardization of optimized mycobacterial interspersed repetitive unit-variable-number tandem repeat typing of Mycobacterium tuberculosis. J.Clin.Microbiol. 44: Brudey, K., J. R. Driscoll, L. Rigouts, W. M. Prodinger, A. Gori, S. A. Al Hajoj, C. Allix, L. Aristimuno, J. Arora, V. Baumanis, L. Binder, P. Cafrune, A. Cataldi, S. Cheong, R. Diel, C. Ellermeier, J. T. Evans, M. Fauville-Dufaux, S. Ferdinand, d. Garcia, V, C. Garzelli, L. Gazzola, H. M. Gomes, M. C. Gutierrez, P. M. Hawkey, P. D. van Helden, G. V. Kadival, B. N. Kreiswirth, K. Kremer, M. Kubin, S. P. Kulkarni, B. Liens, T. Lillebaek, H. M. Ly, C. Martin, C. Martin, I. Mokrousov, O. Narvskaia, Y. F. Ngeow, L. Naumann, S. Niemann, I. Parwati, M. Z. Rahim, V. Rasolofo-Razanamparany, T. Rasolonavalona, M. L. Rossetti, S. Rusch-Gerdes, A. Sajduda, S. Samper, I. Shemyakin, U. B. Singh, A. Somoskovi, R. Skuce, D. van Soolingen, E. M. Streicher, P. N. Suffys, E. Tortoli, T. Tracevska, V. Vincent, T. C. Victor, R. Warren, S. F. Yap, K. Zaman, F. Portaels, N. Rastogi, and C. Sola Mycobacterium tuberculosis complex genetic diversity : mining the fourth international spoligotyping database (SpolDB4) for classification, population genetics and epidemiology. BMC.Microbiol. 6: Demay, C., B. Liens, T. Burguiere, V. Hill, D. Couvin, J. Millet, I. Mokrousov, C. Sola, T. Zozio, and N. Rastogi SITVITWEB--a publicly available international multimarker database for studying Mycobacterium tuberculosis genetic diversity and molecular epidemiology. Infect.Genet.Evol. 2: Hunter, P. R. and M. A. Gaston Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J.Clin.Microbiol. 26: Downloaded from on September 4, 208 by guest Cohen, T. and M. Murray Modeling epidemics of multidrug-resistant M. tuberculosis of heterogeneous fitness. Nat.Med. 0:7-2.

19 Borgdorff, M. W., M. Sebek, R. B. Geskus, K. Kremer, N. Kalisvaart, and S. D. Van. 20. The incubation period distribution of tuberculosis estimated with a molecular epidemiological approach. Int.J.Epidemiol. 40: Gagneux, S., C. D. Long, P. M. Small, T. Van, G. K. Schoolnik, and B. J. Bohannan The competitive cost of antibiotic resistance in Mycobacterium tuberculosis. Science 32: Calver, A. D., A. A. Falmer, M. Murray, O. J. Strauss, E. M. Streicher, M. Hanekom, T. Liversage, M. Masibi, P. D. van Helden, R. M. Warren, and T. C. Victor Emergence of increased resistance and extensively drug-resistant tuberculosis despite treatment adherence, South Africa. Emerg.Infect.Dis. 6: Uys, P. W., R. Warren, P. D. van Helden, M. Murray, and T. C. Victor Potential of rapid diagnosis for controlling drug-susceptible and drug-resistant tuberculosis in communities where Mycobacterium tuberculosis infections are highly prevalent. J.Clin.Microbiol. 47: Perez-Velez, C. M. and B. J. Marais Tuberculosis in children. N.Engl.J.Med. 367: Hesseling, A. C., S. Kim, S. Madhi, S. Nachman, H. S. Schaaf, A. Violari, T. C. Victor, G. McSherry, C. Mitchell, and M. F. Cotton High prevalence of drug resistance amongst HIV-exposed and -infected children in a tuberculosis prevention trial. Int J.Tuberc.Lung Dis. 6: Strauss, O. J., R. M. Warren, A. Jordaan, E. M. Streicher, M. Hanekom, A. A. Falmer, H. Albert, A. Trollip, E. Hoosain, P. D. van Helden, and T. C. Victor Spread of a lowfitness drug-resistant Mycobacterium tuberculosis strain in a setting of high human immunodeficiency virus prevalence. J.Clin.Microbiol. 46: Downloaded from on September 4, 208 by guest

20 Seddon, J. A., A. C. Hesseling, B. J. Marais, A. Jordaan, T. Victor, and H. S. Schaaf The evolving epidemic of drug-resistant tuberculosis among children in Cape Town, South Africa. Int J.Tuberc.Lung Dis. 6: Lawn, S. D., L. G. Bekker, K. Middelkoop, L. Myer, and R. Wood Impact of HIV infection on the epidemiology of tuberculosis in a peri-urban community in South Africa: the need for age-specific interventions. Clin.Infect.Dis. 42: Mlambo, C. K., R. M. Warren, X. Poswa, T. C. Victor, A. G. Duse, and E. Marais Genotypic diversity of extensively drug-resistant tuberculosis (XDR-TB) in South Africa. Int.J.Tuberc.Lung Dis. 2: Chihota, V. N., B. Muller, C. K. Mlambo, M. Pillay, M. Tait, E. M. Streicher, E. Marais, G. D. van der Spuy, M. Hanekom, G. Coetzee, A. Trollip, C. Hayes, M. E. Bosman, N. C. Gey Van Pittius, T. C. Victor, P. D. van Helden, and R. M. Warren Population structure of multi- and extensively drug-resistant Mycobacterium tuberculosis strains in South Africa. J.Clin.Microbiol. 50: Chihota, V., L. Apers, S. Mungofa, W. Kasongo, I. M. Nyoni, R. Tembwe, G. Mbulo, M. Tembo, E. M. Streicher, G. D. van der Spuy, T. C. Victor, H. P. van, and R. M. Warren Predominance of a single genotype of Mycobacterium tuberculosis in regions of Southern Africa. Int J.Tuberc.Lung Dis. : Shah, N. S., J. Richardson, P. Moodley, S. Moodley, P. Babaria, M. Ramtahal, S. K. Heysell, X. Li, A. P. Moll, G. Friedland, A. W. Sturm, and N. R. Gandhi. 20. Increasing drug resistance in extensively drug-resistant tuberculosis, South Africa. Emerg.Infect.Dis. 7: Downloaded from on September 4, 208 by guest Drobniewski, F., Y. Balabanova, V. Nikolayevsky, M. Ruddy, S. Kuznetzov, S. Zakharova, A. Melentyev, and I. Fedorin Drug-resistant tuberculosis, clinical virulence, and the dominance of the Beijing strain family in Russia. JAMA 293:

21 Toungoussova, O. S., P. Sandven, A. O. Mariandyshev, N. I. Nizovtseva, G. Bjune, and D. A. Caugant Spread of drug-resistant Mycobacterium tuberculosis strains of the Beijing genotype in the Archangel Oblast, Russia. J.Clin.Microbiol. 40: Marais, B. J., T. C. Victor, A. C. Hesseling, M. Barnard, A. Jordaan, W. Brittle, H. Reuter, N. Beyers, P. D. van Helden, R. M. Warren, and H. S. Schaaf Beijing and Haarlem genotypes are overrepresented among children with drug-resistant tuberculosis in the Western Cape Province of South Africa. J.Clin.Microbiol. 44: Almeida, D., C. Rodrigues, T. F. Ashavaid, A. Lalvani, Z. F. Udwadia, and A. Mehta High incidence of the Beijing genotype among multidrug-resistant isolates of Mycobacterium tuberculosis in a tertiary care center in Mumbai, India. Clin.Infect.Dis. 40: van der Spuy, G. D., K. Kremer, S. L. Ndabambi, N. Beyers, R. Dunbar, B. J. Marais, P. D. van Helden, and R. M. Warren Changing Mycobacterium tuberculosis population highlights clade-specific pathogenic characteristics. Tuberculosis.(Edinb.) 89: Hanekom, M., N. C. Gey Van Pittius, C. McEvoy, T. C. Victor, P. D. van Helden, and R. M. Warren. 20. Mycobacterium tuberculosis Beijing genotype: a template for success. Tuberculosis.(Edinb.) 9: Abebe, F. and G. Bjune The emergence of Beijing family genotypes of Mycobacterium tuberculosis and low-level protection by bacille Calmette-Guerin (BCG) vaccines: is there a link? Clin.Exp.Immunol. 45: Toungoussova, O. S., D. A. Caugant, P. Sandven, A. O. Mariandyshev, and G. Bjune Impact of drug resistance on fitness of Mycobacterium tuberculosis strains of the W- Beijing genotype. FEMS Immunol.Med.Microbiol. 42: Downloaded from on September 4, 208 by guest

22 Murray, E. J., B. J. Marais, G. Mans, N. Beyers, H. Ayles, P. Godfrey-Faussett, S. Wallman, and V. Bond A multidisciplinary method to map potential tuberculosis transmission 'hot spots' in high-burden communities. Int J.Tuberc.Lung Dis. 3: Marais, B. J., M. C. Raviglione, P. R. Donald, A. D. Harries, A. L. Kritski, S. M. Graham, W. M. El-Sadr, M. Harrington, G. Churchyard, P. Mwaba, I. Sanne, S. H. Kaufmann, C. J. Whitty, R. Atun, and A. Zumla Scale-up of services and research priorities for diagnosis, management, and control of tuberculosis: a call to action. Lancet 375: Downloaded from on September 4, 208 by guest

23 Figure. Phylogenetic links of Beijing family strains using 2-MIRU (A) and 24-MIRU (B) profiles. Solid lines show a single loci-miru change, while dotted lines show 2 (black coloured) or more (grey colored) changes. Circles show 2-loci MIRU international type (MIT) numbers and the color of the circles reflects the number of clinical isolates identified (N=7), illustrating unique (sky-blue) versus clustered isolates (deep blue, 2 to 5 strains; dark blue, 5 to 0 strains; brown, 0 to 20 strains; red, 20 strains and more). Additional color groups demonstrate likely clusters with minimal strain variation. Downloaded from on September 4, 208 by guest

24 TABLES: Table. Demographics and first-line drug resistance profile of MDR-TB patients diagnosed in Johannesburg, South Africa (March Dec 2007) Total Gender Male Age category (years) < >60 Unknown Mean adult age ( 5yrs) Male Female First-line drug resistance profile Isoniazid and rifampicin only Isoniazid, rifampicin and ethambutol Isoniazid, rifampicin, ethambutol and streptomycin *HIV infection status unknown 434 (%) 265 (6.) 33 (7.6) 92 (2.2) 20 (48.4) 63 (4.5) 5 (.2) 3 (7.) 36.9 years 33.8 years 94 (2.7) 02 (23.5) 238 (54.8) Downloaded from on September 4, 208 by guest

25 Table 2. Clustering of predominant MDR strains; 2-MIRU sub-clustering of the 0 largest spoligotype clusters Spoligotype cluster SIT* (Octal code) Clade SIT ( ) Beijing SIT60 ( ) LAM4 SIT53 ( ) T SIT50 ( ) H3 SIT34 ( ) S SIT806 ( ) EAI_SOM SIT92 ( ) X3 SIT 48 ( ) EAI-SOM SIT33 ( ) LAM3 Total strains N=337 (%) 7 (6.4) 69 (5.9) 45 (0.4) 37 (8.5) 32 (7.4) 20 (4.6) 20 (4.6) 7 (3.9) 6 (3.7) 2-MIRU sub-clustering MIT** designation (number of strains per cluster) MIT7 (n=7), MIT84 (n=3), MIT96 (n=5), MIT99 (n=8), MIT0 (n=4), MIT04 (n=), MIT245 (n=4), MIT254 (n=0), MIT202 (n=3), MIT203 (n=2), MIT204 (n=2), MIT205 (n=4), MIT206 (n=2), Orphan (n=6) MIT34 (n=), MIT25 (n=), MIT246 (n=64), MIT593 (n=), MIT99 (n=), orphan (n=), MIT8 (n=9), MIT3 (n=), MIT33 (n=), MIT34 (n=), MIT56 (n=), MIT58 (n=3), MIT224 (n=5), MIT23 (n=), MIT246 (n=2), MIT953 (n=), MIT06 (n=), Orphan (n=9) MIT23 (n=), MIT2(n=35), Orphan (n=) MIT22 (n=6), MIT232 (n=5), MIT256 (n=3), MIT804 (n=), MIT26 (n=8), MIT27 (n=2), MIT28 (n=3), Orphan (n=4) MIT20 (n=8), Orphan (n=2) MIT224 (n=7), Orphan (n=3) MIT64 (n=2), MIT208 (n=8), MIT209 (n=6), Orphan (n=) MIT23 (n=8), MIT236 (n=2), MIT246 (n=), MIT20 (n=), MIT23 (n=), MIT24 (n=), Orphan (n=2) Downloaded from on September 4, 208 by guest 495 SIT 244 ( ) T 0 (2.3) MIT32 (n=2), MIT34 (n=), MIT39 (n=), MIT7 (n=4), MIT6 (n=), MIT289 (n=) *SIT Shared international type **MIT MIRU international type

26 Table 3. Detailed analysis of Beijing family strains; 24-MIRU sub-clustering of 2-MIRU-defined clusters MIRU cluster MIT Number N=7 24-MIRU sub-cluster _ Number N=7 MIT MIT MIT MIT MIT MIT MIT MIT MIT MIT MIT MIT Orphan Downloaded from on September 4, 208 by guest

27 _ _3424_ _ denotes alleles that could not be amplified at particular MIRU loci (missing alleles) Downloaded from on September 4, 208 by guest

28 505 Table 4. Locus-specific discriminatory value of 24-MIRU in Beijing family strains MIRU locus* MIRU alias No of alleles No of repeats HGDI Discrimination 2- MIRU 5- MIRU 24- MIRU 54 MIRU2 2 0 Poor * * 580 MIRU4 2 0 Poor * * * 960 MIRU0 2 2 to 3 0 Poor * * * 644 MIRU6 2 2 to Poor * * * 2059 MIRU Poor * * 253 MIRU Poor * * 2687 MIRU24 0 Poor * * 2996 MIRU to Moderate * * * 3007 MIRU27 5 to Moderate * * 392 MIRU3 2 to High * * * 4348 MIRU to Moderate * * 802 MIRU40 3 to Moderate * * * 955 Mtub2 4 3 to Moderate * * 265 ETR-A 2 3 to Poor * * 263b QUB-b 4 2 to High * * 4052 QUB to 0.4 Moderate * * 424 Mtub to Moderate * * 240 Mtub to Poor * * 456 QUB to Poor * * 3690 Mtub to Poor * * 577 ETR-C 3 3 to Moderate * * 2347 Mtub to Poor * 246 ETR-B 2 0 Poor * 37 Mtub to 4 0 Poor * Downloaded from on September 4, 208 by guest 28

29 A B Downloaded from on September 4, 208 by guest