Extensively Drug-Resistant Mycobacterium tuberculosis in Taiwan.

(1)

M A J O R A R T I C L E

Extensively Drug-Resistant Mycobacterium

tuberculosis during a Trend of Decreasing Drug

Resistance from 2000 through 2006 at a Medical

Center in Taiwan

Chih-Cheng Lai,1_{Che-Kim Tan,}4_{Yu-Tsung Huang,}2,3_{Chien-Hong Chou,}5_{Chien-Ching Hung,}2_{Pan-Chyr Yang,}2 Kwen-Tay Luh,2,3_{and Po-Ren Hsueh}2,3

1_{Department of Internal Medicine, Yi-Min Hospital, and Departments of}2_{Internal Medicine and}3_{Laboratory Medicine, National Taiwan University} Hospital, National Taiwan University College of Medicine, Taipei,4_{Department of Intensive Care Medicine, Chi-Mei Medical Center, Tainan, and} 5_{Department of Internal Medicine, National Taiwan University Hospital Yun-Lin Branch, Yun-Lin, Taiwan}

Background. Drug resistance rates are one of the most important aspects in the national tuberculosis (TB)

control program, and drug-resistant TB, especially extensively drug-resistant (XDR) TB, is not well understood in Taiwan. The objectives of this study were to investigate the prevalence of drug resistance from 2000 through 2006 and to identify XDR TB isolates to elucidate the clinical characteristics of patients with XDR TB at National Taiwan University Hospital.

Methods. The prevalence of drug resistance among clinical, nonduplicate Mycobacterium tuberculosis isolates

was analyzed. Testing of susceptibility to antituberculosis agents, including isoniazid, rifampicin, ethambutol, streptomycin, rifabutin, ofloxacin, ethinamide, and para-aminosalicylic acid, was performed using the proportional method. Minimum inhibitory concentrations of amikacin, capreomycin, isepamycin, linezolid, cycloserine, cip-rofloxacin, levofloxacin, moxifloxacin, and gemifloxacin were determined for 40 available multidrug-resistant M.

tuberculosis isolates.

Results. Significant decreasing trends in rates of resistance to isoniazid, ethambutol, and at least 1 of the 3

first-line agents were observed among 2625 M. tuberculosis isolates from 2000 through 2006. Among these 2625 isolates, 150 (5.7%) were multidrug resistant, and 10 M. tuberculosis isolates (0.4%) fulfilled the definition of XDR

M. tuberculosis. Nine (90%) of 10 patients with XDR TB had a previous history of TB and received anti-TB

treatment before acquisition of XDR TB.

Conclusions. The remaining high prevalence of multidrug-resistant TB and the presence of XDR TB during

a trend of decreasing drug resistance are alarming. Continuous surveillance of clinical isolates of M. tuberculosis is needed to identify XDR TB, especially in patients who have a history of TB and have received prior anti-TB treatment.

The World Health Organization estimates for 2005 in-dicated that the Southeast Asian Region had the largest number of new tuberculosis (TB) cases, which ac-counted for 35% of the global burden of new and re-lapse cases. There were 1.6 million deaths resulting from TB in 2005 [1]. In 2005, the Center for Disease Control

Received 1 May 2008; accepted 30 June 2008; electronically published 20 August 2008.

Reprints or correspondence: Dr. Po-Ren Hsueh, Depts. of Laboratory Medicine and Internal Medicine, National Taiwan University Hospital, No. 7, Chung-Shan South Rd., Taipei, 100, Taiwan ([email protected]).

Clinical Infectious Diseases 2008; 47:e57–63

DOI: 10.1086/591702

of Taiwan recorded 15,378 newly diagnosed TB cases. A TB incidence of 72.7 cases per 100,000 population and a TB mortality rate of 4.2 deaths per 100,000 pop-ulation were reported [2]. Despite being one of the oldest known diseases, TB is still a growing problem worldwide. Drug-resistant TB mainly arises from in-consistent or partial treatment because of poor drug compliance, incorrect treatment regimens, or an un-reliable drug supply. Isoniazid was introduced for the treatment of TB in 1952. The prevalence of isoniazid resistance ranged from 8.4% of isolates in the early 1960s to 22.6% in the 1970s but decreased to 6.8% in the 1980s [3]. Rifampin-based chemotherapy has been widely used in Taiwan since 1978. Although there was

at Taipei Medical University Library on May 21, 2011

cid.oxfordjournals.org

(2)

no rifampin resistance in the early 1980s in Taiwan [4], resis-tance to rifampin gradually increased thereafter. Multidrug-resistant (MDR) TB is defined as TB with resistance to both isoniazid and rifampin, the 2 most effective anti-TB drugs. It is a particularly dangerous form of drug-resistant TB that has resulted from inappropriate treatment in Taiwan and represents a growing threat. Inappropriate treatment for drug-resistant TB not only results in treatment failure but is also responsible for further dissemination of drug-resistant strains, rendering the control of TB a more difficult public health issue. Furthermore, extensively drug-resistant (XDR) TB, defined as TB that is re-sistant to at least isoniazid and rifampin (MDR-TB), in addition to any fluoroquinolone, and⭓1 of the 3 injectable drugs (ca-preomycin, kanamycin, and amikacin), has recently emerged as a global health problem, threatening the success of TB-con-trol programs worldwide [5, 6].

In Taiwan, there are only 8 hospitals and 1 official TB-control institute that have routinely performed antimycobacterial sus-ceptibility testing for clinical isolates obtained from individuals with TB in recent decades. However, the use of different meth-ods for susceptibility testing and different definitions of resis-tance to isoniazid have contributed to variations in reported resistance rates [7]. Moreover, rates of resistance to second-line agents have rarely been reported in Taiwan. The objective of this study was to investigate the prevalence of drug resistance in clinical and nonduplicate isolates of M. tuberculosis from 2000 through 2006 at National Taiwan University Hospital (NTUH; Taipei, Taiwan). We also try to identify XDR M.

tu-berculosis isolates to elucidate the clinical characteristics of

pa-tients with XDR TB.

PATIENTS AND METHODS

Setting and bacterial isolates. This study was conducted at

NTUH, a 2000-bed tertiary care center in northern Taiwan. Isolates obtained from patients who had a culture positive for

M. tuberculosis at NTUH from January 2000 through December

2006 were included in this retrospective analysis. A total of 2625 nonduplicate isolates from 2625 patients were collected during the 7-year period. These isolates were recovered from various clinical specimens, including 2253 (85.8%) from re-spiratory secretions (sputum and bronchial washing), 190 (7.2%) from pleural effusion specimens, 73 (2.8%) from sur-gical wounds samples, 31 (1.2%) from lymph node specimens, 10 (0.4%) from pericardial effusion specimens, and the rest from other specimens. Nonduplicate isolates were defined as a single isolate collected for evaluation from a single patient who visited the hospital. If a patient had multiple isolates, only the first isolate was analyzed. All specimens were processed and pretreated as described elsewhere [8, 9]. A fluorometric BAC-TEC technique (BACBAC-TEC MGIT 960 system; Becton-Dickinson Diagnostic Instrument Systems) was used for routine culture.

Drug susceptibility testing. Testing of susceptibility to

first-line anti-TB drugs, including isoniazid (0.2 mg/mL and 1.0 mg/ mL), rifampin (1 mg/mL), and ethambutol (5 mg/mL), was performed in the mycobacteriology laboratory of NTUH. Since 1 January 2005, testing of susceptibility to second-line anti-TB drugs, including streptomycin (2 mg/mL and 10 mg/mL), ri-fabutin (0.5 mg/mL), ofloxacin (1 mg/mL), ethionamide (5 mg/ mL), and para-aminosalicyclic acid (2 mg/mL), was also per-formed. Drug susceptibility testing for these anti-TB drugs was performed in the mycobacteriology laboratory of NTUH using the agar proportion method [10]. M. tuberculosis suspension was inoculated onto Middlebrook 7H10 agar (BBL Microbi-ology Systems) that contained anti-TB drugs at respective con-centrations. The number of colony-forming units growing on the drug-containing medium was compared with the number of colony-forming units growing on a drug-free medium. Iso-lates for which growth on the drug-containing media presented

!_{1% of the number of colonies that developed on the}

drug-free media were considered to be resistant to that agent. For quality control, the standard sensitive strain, H37Rv, and the resistant strain, Vertulo, were also tested for drug susceptibility with the same procedures.

Drug resistance was defined as resistance to isoniazid (0.2 mg/mL), rifampin (1 mg/mL), ethambutol (5 mg/mL), or strep-tomycin (2 mg/mL). An MDR isolate was defined as being resistant to at least isoniazid (0.2 mg/mL) and rifampin (1 mg/ mL). XDR M. tuberculosis was defined as resistant to at least isoniazid and rifampin, as well as resistant to any fluoroquin-olone and⭓1 of the 3 injectable drugs (capreomycin, kana-mycin, and amikacin) [5].

HIV-infection status and drug resistance. Among the 2625

patients, 504 patients had received antibody screening and/or Western blot confirmation tests for HIV. For detecting HIV-1 and/or HIV-2 antibody, a passive particle agglutination method (Bio-Rad) was used through 2006 and an ELISA method (Axsym; Abbott) was used in 2007 and after. For confirmation testing, an immunoblotting method (Bio-Rad) was used during the study period. Patients who had both positive antibody screening results and immunoblotting test results positive for HIV were considered to be HIV infected. Patients with results negative for HIV antibody were not considered to be infected with HIV. Resistance profiles of isolates collected from these patients were analyzed on the basis of HIV status of the patient.

Determination of MICs. MICs of 9 second-line anti-TB

agents for 40 preserved MDR M. tuberculosis isolates recovered during the period 2000–2006 were determined using the agar dilution method. Concentrations of 0.03–32 mg/mL were tested for amikacin, capreomycin, ciprofloxacin, levofloxacin, moxi-floxacin, gemimoxi-floxacin, linezolid, cycloserine, and isepamicin. The MICs were determined by serial dilution on agar plates as described elsewhere [10]. The MIC for each isolate-drug pair

(3)

Figure 1. Trends of rates of resistance to isoniazid, ethambutol, rif-ampin, and any 1 of these 3 drugs (A) and multidrug resistance (B) among

Mycobacterium tuberculosis isolates recovered from patients treated at

the National Taiwan University Hospital (Taipei, Taiwan) from 2000 through 2006, determined using the modified proportional method. P values!_.05

were considered to be statistically significant.

Table 1. Drug resistance patterns of second-line agents for mul-tidrug-resistant (MDR) and non-MDR Mycobacterium tuberculosis isolates obtained from 2005 through 2006, determined using the modified proportional method.

Agent, concentration No. (%) of resistant isolates P Non-MDR isolates (n p 920) MDR isolates (n p 42) Streptomycin 2 mg/mL 61 (6.6) 22 (52.4) !.001 10 mg/mL 24 (2.6) 27 (64.3) !.001 Rifabutin, 0.5 mg/mL 2 (0.2) 20 (47.6) !.001 Ofloxacin, 2 mg/mL 1 (0.1) 7 (16.7) !.001 Ethionamide, 5 mg/mL 4 (0.4) 10 (23.8) !.001 Para-aminosalicylic acid, 2 mg/mL 10 (1.1) 7 (16.7) !.001

was defined as the lowest concentration of the agent that in-hibited1_{99% of the growth of colonies on the drug-free control}

culture. Resistance was presumptively defined as follows: MICs of1_{2.5 mg/mL for capreomycin;} 1_{2 mg/mL for ciprofloxacin;} 1_{1 mg/mL for levofloxacin, linezolid, amikacin, and isepamicin;}

and1_{0.5 mg/mL for moxifloxacin [11–14].}

Statistical analysis. Differences in drug susceptibility

be-tween MDR M. tuberculosis and non-MDR M. tuberculosis iso-lates and between isoiso-lates obtained from HIV-infected and from non–HIV-infected patients were analyzed using the x2 _test.

Drug resistance trends over time were evaluated by Cochran-Armitage trend test. A P value of!_{.05 was considered to be}

statistically significant. RESULTS

A total of 2625 nonduplicate M. tuberculosis isolates were col-lected during the study period. Of these isolates, 403 (15.4%) were resistant to isoniazid, 175 (6.7%) were resistant to rif-ampin, 224 (8.5%) were resistant to ethambutol, and 613 (23.4%) were resistant to any 1 of these 3 drugs. A total of 150 isolates (5.7%) met the criteria for classification as MDR M.

tuberculosis. Trend analysis showed that the resistance rate to

isoniazid, to ethambutol, and to any 1 of isoniazid, ethambutol, and rifampin increased significantly during the 7-year study period (figure 1).

Additional tests for susceptibility to 5 second-line anti-TB agents, including streptomycin, rifabutin, ofloxacin, ethion-amide, and para-aminosalicyclic acid, were performed for 962 isolates in 2005 and 2006. Of these isolates, 42 were MDR M.

tuberculosis. The rate of resistance to each of the 5 agents was

significantly higher for MDR isolates than it was for non-MDR isolates (table 1).

The MICs at which 50% of the isolates were inhibited (MIC50) and at which 90% of the isolates were inhibited (MIC90)

and the MIC ranges for the 40 MDR M. tuberculosis isolates are shown in table 2. Among the 4 fluoroquinolones tested, moxifloxacin showed the greatest activity against the MDR M.

tuberculosis isolates, followed by levofloxacin, and ciprofloxacin.

Gemifloxacin was the most inactive fluoroquinolone against the isolates tested. Of the other 5 agents, linezolid and isepam-icin were most active against MDR M. tuberculosis isolates, followed by cycloserine, capreomycin, and amikacin.

Demographic characteristics and clinical manifestations of the 10 patients with XDR TB are shown in table 3. Of the 10 XDR M. tuberculosis isolates, all were resistant to ofloxacin and levofloxacin, 1 (10%) was susceptible to ciprofloxacin, and 1 (10%) was susceptible to moxifloxain. Most of the patients were male, and the mean age (SD) of the patients infected with XDR M. tuberculosis was56.8 16.6 years. Diabetes mellitus was the most frequent underlying disease (found in 60% of patients), followed by chronic pulmonary disease (20%), lung cancer (10%), and end-stage renal disease (10%). A total of 90% of patients had a history of TB, and 50% of patients had received fluoroquinolones1_{1 month before acquisition of XDR}

(4)

Table 2. In vitro activity of 8 agents against 40 multidrug-resis-tant Mycobacterium tuberculosis isolates.

Agent

MIC, mg/mL

Range MIC50 MIC90

Amikacin 0.25 to132 1 132 Isepamicin !0.03 to 1 0.5 1 Capreomycin 2 to132 4 32 Gemifloxacin 0.5 to132 16 132 Ciprofloxacin 0.25 to132 1 16 Levofloxacin 0.25 to 16 0.5 8 Moxifloxacin 0.25 to 8 0.25 8 Linezolid !0.03 to 4 0.5 0.5 Cycloserine !0.03 to 32 1 16

Table 3. Demographic and clinical features of 10 patients with extensively drug-resistant (XDR) tuberculosis (TB) infection.

Patient Year Age, years Sex Underlying disease Acid-fast stain result Previous history of TB Radiographic findings of cavitary lesions Treatment with anti-TB drugs⭓1 month before acquisition of XDR TB

1 2000 25 F … Negative Yes Yes H, E, R, Z, levofloxacin, streptomycin

2 2004 73 M DM Negative Yes Yes H, E, R, Z, levofloxacin

3 2004 82 M COPD Negative Yes No NA

4 2004 53 M DM Positive No Yes H, E, R, Z

5 2005 59 M DM Positive Yes No H, E, R, Z, streptomycin

6 2005 49 M DM, ESRD, HCC s/p

transplant

Negative Yes Yes No

7 2005 59 F DM Positive Yes Yes H, E, R, Z, moxifloxacin

8 2006 65 F Lung cancer Negative Yes No H, E, R, Z

9 2006 65 M DM, pneumoconiosis Positive Yes Yes H, E, R, Z, moxifloxacin,

streptomy-cin, levofloxastreptomy-cin, PAS, ethionamide

10 2007 38 M No Positive Yes Yes H, E, R, Z, streptomycin,

levofloxa-cin, amikalevofloxa-cin, PAS

NOTE. COPD, chronic obstructive pulmonary disease; DM, diabetes mellitus; E, ethambutol; ESRD, end stage renal disease; H, isoniazid; HCC, hepa-tocellular carcinoma; NA, not applicable; PAS, para-aminosalycilic acid; R. rifampin; s/p, status post; Z, pyrazinamide.

TB. In 7 patients, radiological findings showed cavitary lesions, but only 1 patient had pleural effusion.

Of the 504 patients with TB for whom data regarding HIV infection status were available, 75 were HIV infected, and 429 were not HIV infected. There were no significant differences between the 2 groups with respect to most of the drug resistance patterns, except for a significantly higher prevalence of high-level isoniazid resistance (P p .03) among the HIV-infected patients (table 4).

DISCUSSION

In 2005, there were an estimated 8.8 million new cases of TB and 1.6 million TB-related deaths worldwide [1]. Control of TB remains one of the most challenging issues in global health [1]. A new and potentially devastating threat to TB control is the emergence of strains that cannot be cured by standard anti-TB drug regimens. Drug resistance rates are regarded as one of the most important aspects of surveillance in the national TB control program in Taiwan.

In this study, the overall rate of resistance to any 1 of the 3 drugs isoniazid, rifampin, or ethambutol was 23.4%. These rates are lower than those from other regions, including south-ern Taiwan (29%) [15], Guatemala (30%) [16], and New York (31%) [17]. Liaw et al. [18] reported that, during the period 1998–2002, 19.0% of TB isolates analyzed at NTUH were re-sistant to isoniazid, 6.1% were rere-sistant to rifampin, and 15.7% were resistant to ethambutol. Our study revealed a decrease in the rates of resistance to isoniazid (from 16.7% to 12.4%) and ethambutol (from 9.1% to 2.5%) in the 2003–2006 period. In fact, this study found decreasing rates of resistance to isoniazid, ethambutol, and any 1 of the 3 drugs isoniazid, rifampin, and ethambutol during the 2000–2006 period. Similar decreasing rates of resistance have been reported by recent studies from Taiwan [19, 20], Hong Kong [21], and Saudi Arabia [22].

In Taiwan, the implementation of 2 effective interventions might explain the decreasing rates of resistance to anti-TB drugs. In 1997, stricter regulation mandated that each treated TB case be reported to the Center for Disease Control of Tai-wan. Since then, the percentage of patients with TB who receive a complete course of treatment has increased, and the per-centage of those lost to follow-up has decreased. Second, di-rectly observed short-course therapy, which is a proven and effective measure, was also started in Taiwan during this period. Our findings suggest that these measures have increased the rate of treatment completion and might have played a role in decreasing the emergence and spread of drug-resistant TB.

In spite of the encouraging findings of decreasing rates of resistance to anti-TB agents, MDR TB still poses a challenge to TB control. In this study, 3.0%–7.7% of the isolates were MDR M. tuberculosis, and this percentage remained fairly stable during the study period. This prevalence is considerably higher than the median rate of MDR M. tuberculosis (1.0%; range,

(5)

Table 4. Comparison of drug resistance for Mycobacterium

tu-berculosis isolates recovered from 504 patients for whom HIV

infection status data were available from 2000 through 2006, de-termined using the modified proportional method.

Agent, concentration No. (%) of isolates P From HIV-infected patients (n p 75) From non–HIV-infected patients (n p 429) Isoniazid 0.2 mg/mL 13 (17.3) 53 (12.4) .24 1.0 mg/mL 9 (12) 23 (5.4) .03a Rifampin, 1 mg/mL 5 (6.7) 17 (4.0) .29 Ethambutol, 5 mg/mL 3 (4.0) 11 (2.6) .49 Streptomycin 2 mg/mL 5 (6.7) 39 (9.1) .49 10 mg/mL 2 (2.7) 24 (5.6) .29

Resistant to any drug 19 (25.3) 76 (17.7) .12 Multidrug resistant 2 (2.7) 14 (3.3) .79

a

Statistically significant.

0.0%– 14.2%) in the 76 countries or geographical settings in-cluded in the World Health Organization/International Union Against Tuberculosis and Lung Disease surveillance report for 1999–2002 [23]. However, comparison of MDR TB prevalence in an individual country with prevalence in a referral hospital is inappropriate, because the referral hospital receives the most complicated cases.

Previous studies from Taiwan have reported a prevalence of MDR TB of 5.1%–17.3% [15, 18, 19, 24, 25]. Moreover, a high percentage of resistance to the second-line anti-TB agents usu-ally used to treat MDR TB was also noted [15, 18, 19, 24, 25]. The present study clearly demonstrated that there is higher rate of resistance to streptomycin, rifabutin, ofloxacin, ethionamide, and para-aminosalicyclic acid among MDR isolates than among non-MDR isolates, with overall rates of resistance to these 5 agents ranging from 16.7% to 52.4%. The high prevalence of MDR TB and the high rate of resistance to both first-line and second-line agents is still a growing threat in Taiwan, and more-effective TB-control interventions and more-potent anti-TB agents are urgently needed.

The recent emergence of XDR TB has become another global health problem that constitutes a deadly threat to patients and hampers TB-control programs [6]. In Taiwan, XDR TB has rarely been reported, and only 22 (10.2%) of 215 MDR isolates have fulfilled the criteria for XDR TB [26]. Although only 10 isolates of XDR TB were identified in our study, this low num-ber was attributed to a failure to perform drug susceptibility testing for injectable drugs and fluoroquinolone for all MDR

M. tuberculosis isolates. Our study revealed that patients with

XDR TB had a high prevalence of previous TB and that many had received prior anti-TB treatment.

These findings are consistent with those of a previous study

from Korea [27]. Fluoroquinolones and aminoglycosides were prescribed to 5 and 4 patients, respectively. The rate of treat-ment with second-line anti-TB drugs, such as fluoroquinolones and aminoglycosides, was lower in our study than in the study by Kim et al. [27], who reported that 35 (81.4%) of 43 and 38 (88.4%) of 43 patients with XDR TB had received fluoroquin-olones and aminoglycosides, respectively. Although the number of cases in our study is limited, our findings suggest the need for continuous surveillance of clinical isolates of M. tuberculosis to identify cases of XDR TB, especially among patients with a previous history of TB and those who have received prior anti-TB treatment, including fluoroquinolones and amino-glycosides.

Kim et al. [28] reported that 37 (86%) of 43 patients with XDR TB had chest radiograph findings showing a cavitary le-sion, but only 2 (4.7%) had diabetes mellitus. Our study re-vealed that patients with XDR TB had a high prevalence of diabetes mellitus and cavitary lesions in the lungs; in addition, men were more likely than women to have XDR TB. These findings may imply that individuals with XDR TB were more likely than others to have pulmonary cavities, but more epi-demiological data is required to clarify the relationship between diabetes mellitus, sex, and XDR TB.

Fluoroquinolones have broad-spectrum antimicrobial activ-ity and may play useful roles in prophylactic treatment for patients exposed to MDR TB, treatment of proven MDR TB, and empirical treatment of TB disease in settings with high rates of MDR TB [29–31]. In this study, we compared the activity of the different fluoroquinolones against 40 clinical isolates of MDR M. tuberculosis. Levofloxacin and moxifloxacin showed better in vitro activity against MDR M. tuberculosis than did other drugs, suggesting their increasingly important role in the treatment of MDR TB. Gemifloxacin had the poorest in vitro activity, not only against MDR M. tuberculosis isolates, but also against non-MDR M. tuberculosis isolates (data not shown). The naphthyridone structure of gemifloxacin was iden-tified as a negative factor in a quantitative structure-activity relationship study of antimycobacterial activity [31], which might explain its poor anti-TB activity.

The activities of other classes of antimicrobials in addition to fluoroquinolones, such as aminoglycosides and oxazolidi-nonoes, were also tested against MDR M. tuberculosis in this study. Our results showed that linezolid displayed potent ac-tivity against MDR TB. These results are consistent with the findings of Tato et al. [32]. Because clinical experience with and in vitro study of linezolid has been limited, its potential role as a treatment for MDR TB deserves additionalevaluation. Our results showed that, among the aminoglycosides tested, isepamicin was the most active antimycobacterial agent against MDR M. tuberculosis. However, an in vivo study in mice found that amikacin was more active than isepamicin against TB [11].

(6)

The reason for these different results remains to be determined, but this difference could be attributable to differences in study design, including the use of an in vitro versus an animal model and the use of different strains of M. tuberculosis versus MDR

M. tuberculosis. Considerable additional study is needed to

eval-uate the potential role of aminoglycosides in the treatment of TB.

Infection with HIV is an important risk factor for the de-velopment of TB. Taiwan has a low prevalence of HIV infection. HIV-positive patients with TB comprise only a small portion of all TB patients in Taiwan. In this study, the rates of drug resistance among isolates from HIV-infected patients were not significantly different from those among isolates from HIV-negative patients. These findings are consistent with a previous study from this institution [14] and the study of Espinal et al. [33], which supported a lack of association between HIV in-fection and the development of MDR TB, per se.

This retrospective and laboratory-based surveillance study had 2 noteworthy limitations. First, we were unable to precisely distinguish between newly diagnosed and previously treated cases and, therefore, were only able to report the combined resistance rate. Second, this study was conducted in a tertiary care center and, as such, its findings might not reflect the overall situation in Taiwan.

In conclusion, although there was a decreasing overall trend of anti-TB drug resistance in recent years, the prevalence of MDR TB remains high, and the presence of XDR TB will im-pose a new challenge in the control of TB. Continuous sur-veillance of clinical isolates of M. tuberculosis is needed to iden-tify MDR TB or XDR TB, especially in patients with a history of TB and those who have received prior anti-TB treatment.

Acknowledgments

Financial support. Institute for Biotechnology and Medicine Industry (DOH97-DC-1501).

Potential conflicts of interest. All authors: no conflicts.

References

1. Global tuberculosis control: surveillance, planning, financing. WHO report 2007. WHO/HTM/TB/2007.376. Geneva: World Health Orga-nization, 2007.

2. Centers for Disease Control of Taiwan. Statistics of communicable diseases and surveillance report, Republic of China, 2005. Available at: http://www.cdc.gov.tw/public/Data/83241741971.pdf. Accessed 18 March 2008.

3. Lee JJ, Lee CN, Suo J, et al. Drug resistance of Mycobacterium

tuber-culosis in eastern Taiwan. Tzu Chi Med J 2003; 15:229–33.

4. Chang CE. Trends of primary drug resistance of Mycobacterium

tu-berculosis isolated from tutu-berculosis patients in the recent 10 years in

Taiwan. In: Taiwan Provincial Health Department. Collected reports of medical research. 1985:137–47.

5. Centers for Disease Control and Prevention. Notice to readers: revised definition of extensively drug-resistant tuberculosis. MMWR Morb Mortal Wkly Rep 2006; 55:1176.

6. Centers for Disease Control and Prevention. Emergence of

Mycobac-terium tuberculosis with extensive resistance to second-line drugs

world-wide, 2000–2004. MMWR Morb Mortal Wkly Rep 2006; 55:301–5. 7. Hsueh PR, Liu YC, So J, Liu CY, Yang PC, Luh KT. Mycobacterium

tuberculosis in Taiwan. J Infect 2006; 52:77–85.

8. Wang JY, Lee LN, Chou CS, et al. Performance assessment of a nested-PCR assay (the RAPID BAP-MTB) and the BD ProbeTec ET system for detection of Mycobacterium tuberculosis in clinical specimens. J Clin Microbiol 2004; 42:4599–603.

9. Wang JY, Lee LN, Hsueh PR, et al. Tuberculous myositis: a rare but existing clinical entity. Rheumatology (Oxford) 2003; 42:836–40. 10. National Committee of Clinical Laboratory Standards. Susceptibility

testing of mycobacteria, Nocardiae, and other aerobic Actinomycetes: approved standards. Wayne, PA: NCCLS, 2003.

11. Lounis N, Ji B, Truffot-Pernot C, Grosset J. Which aminoglycoside or fluoroquinolone is more active against Mycobacterium tuberculosis in mice? Antimicrob Agents Chemother 1997; 41:607–10.

12. Ginsburg AS, Hooper N, Parrish N, et al. Fluoroquinolone resistance in patients with newly diagnosis tuberculosis. Clin Infect Dis 2003; 37: 1448–52.

13. Ginsburg AS, Grosset JH, Bishai WR. Fluoroquinolones, tuberculosis, and resistance. Lancet Infect Dis 2003; 3:432–42.

14. Rusch-Gerdes S, Pfyffer GE, Casal M, Chadwick M, Siddiqi S. Mul-ticenter laboratory validation of the BACTEC MGIT 960 technique for testing susceptibilities of Mycobacterium tuberculosis to classical second-line drugs and new antimicrobials. J Clin Microbiol 2006; 44:688–92. 15. Tsao TC, Chiou W, Lin H, et al. Change in demographic picture and increase of drug resistance in pulmonary tuberculosis in a 10-year interval in Taiwan. Infect 2002; 30:75–80.

16. Harrow EM, Rangel JM, Arriega JM, et al. Epidemiology and clinical consequences of drug-resistant tuberculosis in a Guatemalan hospital. Chest 1998; 113:1452–8.

17. Chawla PK, Klapper PJ, Kamholz SL, Pollack AH, Heurich AE. Drug-resistant tuberculosis in an urban population including patients at risk for human immunodeficiency virus infection. Am Rev Respir Dis

1992; 146:280–4.

18. Liaw YS, Hsueh PR, Yu CJ, Wang SK, Yang PR, Luh KT. Drug resistance pattern of Mycobacterium tuberculosis in a university hospital in Taiwan, 1998–2002. J Formos Med Assoc 2004; 103:671–7.

19. Lu PL, Lee YW, Peng CF, et al. The decline of high drug resistance rate of pulmonary Mycobacterium tuberculosis isolates from a southern Taiwan medical centre, 1996–2000. Int J Antimicrob Agents 2003; 21: 239–43.

20. Wang PD, Lin RS. Drug-resistant tuberculosis in Taipei, 1996–1999. Am J Infect Control 2001; 29:41–7.

21. Kam KM, Yip CW. Surveillance of Mycobacterium tuberculosis drug resistance in Hong Kong, 1986–1999, after the implementation of di-rectly observed treatment. Int J Tuberc Lung Dis 2001; 5:815–23. 22. Al-Tawfiq JA, Al-Muraikhy AA, Abed MS. Susceptibility pattern and

epidemiology of Mycobacterium tuberculosis in a Saudi Arabian hos-pital: a 15-year study from 1989 to 2003. Chest 2005; 128:3229–32. 23. Aziz MA, Wright A, Laszlo A, et al. Epidemiology of antituberculosis

drug resistance (the Global Project on Anti-tuberculosis Drug Resis-tance Surveillance): an updated analysis. Lancet 2006; 368:2142–54. 24. Chiang IH, Yu MC, Bai KJ, et al. Drug resistance patterns of

tuber-culosis in Taiwan. J Formos Med Assoc 1998; 97:581–3.

25. Wang PD. Epidemiology and control of tuberculosis in Taipei. J Infect

2002; 45:82–7.

26. Yu, MC, Wu HN, Jou R. Extensively drug-resistant tuberculosis, Tai-wan. Emerg Infect Dis 2008; 14:849–50.

27. Jeon CY, Hwang SS, Min JH, et al. Extensively drug-resistant tuber-culosis in South Korea: risk factors and treatment outcomes among patients at a tertiary referral hospital. Clin Infect Dis 2008; 46:42–9. 28. Kim HY, Hwang SS, Kim HJ, et al. Impact of extensive drug resistance

on treatment outcomes in non–HIV-infected patients with multidrug-resistant tuberculosis. Clin Infect Dis 2007; 45:1290–5.

29. Gillespie SH, Kennedy N. Fluoroquinolones: a new treatment for tu-berculosis? Int J Tuberc Lung Dis 1998; 2:265–71.

(7)

30. Berning SE. The role of fluoroquinolones in tuberculosis today. Drugs

2001; 61:9–18.

31. Jacobs MR. Activity of quinolones against mycobacteria. Drugs

1995; 49(Suppl 2):67–75.

32. Tato M, de la Pedrosa EG, Canto´n R, et al. In vitro activity of linezolid against Mycobacterium tuberculosis complex, including

multidrug-re-sistant Mycobacterium bovis isolates. Int J Antimicrob Agents 2006; 28: 75–8.

33. Espinal MA, Laserson K, Camacho M, et al. Determinants of drug-resistant tuberculosis: analysis of 11 countries. Int J Tuberc Lung Dis

2001; 5:887–93.