Extensively Drug-Resistant Tuberculosis, Pakistan

Frequency of extensively drug-resistant tuberculosis in Pakistan increased from 1.5% in 2006 to 4.5% in 2009 (p<0.01). To understand the epidemiology, we genotyped selected strains by using spoligotyping, mycobacterial interspersed repetitive units–variable number of tandem repeats, and IS6110 restriction fragment length polymorphism analysis.


The Study
The Aga Khan University Hospital laboratory receives specimens from >180 collection units located in major cities and towns in Pakistan. Specimens for TB cultures are requested by physicians as required and received through passive specimen collection. Past treatment history is usually not available and thus could not be included in this study. Mycobacterium tuberculosis was isolated from the specimens by using Lowenstein-Jensen and MGIT (Becton Dickinson, Franklin Lakes, NJ, USA) media. It was then identifi ed by using the BACTEC NAP TB differentiation test (Becton Dickinson), growth in para-nitrobenzoic acid-containing media, nitrate reduction, and niacin accumulation (4).
MDR TB was defi ned as resistance to at least isoniazid and rifampicin. XDR TB was defi ned as resistance to quinolones and to 1 of the injectable second-line drugs in addition to MDR. Randomly selected isolates were confi rmed as XDR TB by using the GenoType MTBDRsl assay (HAIN Lifesciences, Nehren, Germany). XDR TB isolates were stored at -80°C. Stored XDR TB strains were revived for genotyping; 57 isolates that grew were used for spoligotyping and MIRU-VNTR typing ( DNA was extracted by using the cetyltrimethylammonium bromide method (6). Spoligotyping was performed by using a commercially available kit provided by Isogen Life Science (De Meern, the Netherlands). Spoligotyping based on the 43 spacers of the DR region of M. tuberculosis complex was performed by using primers DRa 5′-GGTTTTGGGTCTGACGAC-3′ and DRb 5′-CCGGAGAGGGGACGGAAAC-3′ as described (7). Negative and positive controls, including template-free PCR-amplifi ed reaction mixture and M. tuberculosis H37Rv DNA, were used with each spoligotype blot.
Data extracted from the computerized information system of the hospital were transferred to the statistical software SPSS version 14.0 (SPSS, Chicago, IL, USA). Frequencies with percentages were computed for each year. We used χ 2 for trend analysis to assess resistance trends over the study period. A p value <5% was considered signifi cant.
The spoligotyping results were entered in the BioNumerics Software version 3.5 (Applied Maths Program, Biosystematica, Ceredigion, UK). A dendrogram was gen-  (8), which includes >40,000 isolates split into 1,030 shared types and >3,530 orphan profi les. The isolates were genotyped by PCR amplifi cation of 15 MIRU-VNTR loci by using standard methods as described (9). Sizes of the PCR fragments and assignment of the various VNTR alleles were also determined by using standard protocol for gel electrophoresis (www.genoscreen.com). All reactions were performed in duplicate by using standard positive and negative controls supplied in the MIRU-VNTR validation kit.
IS6110 RFLP of M. tuberculosis strains were performed by standardized methods (10). Briefl y, XDR TB strains were recovered on LJ medium. DNA was extracted from the strains by standard methods (6,10). PvuII-digested DNA was subjected to agarose gel electrophoresis and Southern blotting. DNA fi ngerprinting was performed by hybridization with the IS6110 by using the enhanced chemiluminescence method (Amersham Biosciences, Piscataway, NJ, USA).
During 2006-2009, a total of 9,523 M. tuberculosis strains were isolated, including 3,682 (38.7%) MDR TB strains. Although the MDR TB rate remained constant (Table 1), the XDR TB rate (expressed as a percentage of MDR TB isolated in a year) showed a signifi cant increase. The XDR TB strains were from specimens received from 23 cities; the largest numbers were from Karachi (22), Hyderabad (13), and Peshawar (12). Mean ± SD age of patients with XDR TB was 37 ± 14 (range 16-80) years; 57.15% were men and 42.85% women.

Conclusions
This study demonstrated a rising XDR-TB trend in Pakistan and raises concerns despite the fact that Pakistan's 2009 XDR rate (4.5%) of MDR TB is below the global av- Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 16, No. 9, September 2010 erage (6.6%-23.7%) (11). Genotyping data are comparable with those from earlier studies (12,13), suggesting dominance of CAS1 strains. The fact that Beijing family strains were 9% of XDR TB isolates vs. 3% prevalence in the overall MDR TB reported in this population (13) suggests that Beijing strains are associated with drug resistance in Pakistan and adjacent countries (14). Furthermore, the fi nding that the XDR TB strains in our study were genetically diverse argues against dissemination of 1 particular genogroup responsible for drug resistance and supports the concept that XDR TB in Pakistanis is likely to be a consequence of inadequate treatment of TB. The challenging sociopolitical situation in Pakistan is likely to exacerbate this public health problem. Emergency measures are required to avoid an exponential rise in drugresistant TB in the country and the region. We recommend that increased XDR TB rates in this area be considered not just of national concern but also be recognized as a regional public health issue requiring introduction of cooperative and support measures aimed at limiting the spread of drugresistant TB within southern Asia.