Extensively Drug-Resistant Tuberculosis, Burkina Faso

Because data from countries in Africa are limited, we measured the proportion of extensively drug-resistant (XDR) tuberculosis (TB) cases among TB patients in Burkina Faso for whom retreatment was failing. Of 34 patients with multidrug-resistant TB, 2 had an XDR TB strain. Second-line TB drugs should be strictly controlled to prevent further XDR TB increase.

Because data from countries in Africa are limited, we measured the proportion of extensively drug-resistant (XDR) tuberculosis (TB) cases among TB patients in Burkina Faso for whom retreatment was failing. Of 34 patients with multidrug-resistant TB, 2 had an XDR TB strain. Second-line TB drugs should be strictly controlled to prevent further XDR TB increase. E xtensively drug-resistant (XDR) tuberculosis (TB) represents an emerging public health problem worldwide, characterized by alarming fatality rates regardless of patients' HIV status (1,2). XDR TB is defi ned as in vitro resistance to isoniazid and rifampin plus any fl uoroquinolone and at least 1 injectable drug (capreomycin, kanamycin, or amikacin).
Since 2006, a total of 49 countries have reported XDR TB (3). Data from countries in Africa are scant, with the exception of South Africa, where the prevalence is high, especially among HIV-infected persons (3). However, these data may refl ect diagnostic limitations rather than the true epidemiologic situation. In Burkina Faso, a low-income country in western Africa, TB prevalence is 226 new cases per 100,000 population (4), and multidrugresistant (MDR) TB is estimated at 2.1% among new patients with smear-positive TB (5). During a systematic search for MDR TB among TB patients in Burkina Faso for whom a retreatment regimen was failing, we documented 2 cases of XDR TB.

The Study
In 2006, a program was established by the National Tuberculosis Program, Burkina Faso; the University of Brescia, Italy; and the San Raffaele Scientifi c Institute, Italy, to perform drug-susceptibility testing (DST) on sputum samples from patients listed in the Burkina Faso Chronic TB Register. To be included in this register, patients must have experienced treatment failure after standard and retreatment regimens consisting of the following (in order): 2 months of streptomycin, rifampin, isoniazid, pyrazinamide, and ethambutol; then 1 month of rifampin, isoniazid, pyrazinamide, and ethambutol; and then 5 months of rifampin, isoniazid, and ethambutol. Failure was defi ned as a persistently positive sputum smear after 5 months of treatment. All patients were informed and consented to the study.
From January 2006 through March 2009, a total of 156 patients with chronic TB were registered in Burkina Faso. For 88 patients, sputum samples were collected before treatment with second-line drugs; for 48, they were collected at 1-12 months of treatment with second-line drugs. Samples were immediately frozen and stored at -20°C before being transferred on dry ice for culture and fi rst-line DST at the University of Brescia and for second-line DST and genotyping at San Raffaele Scientifi c Institute. Samples were cultured on an MGIT 960 automated system (Becton Dickinson Microbiology Systems, Cockeysville, MD, USA) according to the manufacturer's instructions. DST to fi rst-and second-line drugs was performed on all Mycobacterium tuberculosis isolates by classic dilution method.
M. tuberculosis was isolated from 50 samples; 45 patients had not yet taken second-line drugs and 5 had. Of the 50 isolates, 34 (68%) were confi rmed as MDR TB, 29 from untreated patients and 5 from treated patients. We identifi ed 2 cases of XDR TB (5.9% of all MDR TB cases): 1 patient was sampled at month 24 of treatment with secondline drugs, and 1 was initially classifi ed as having MDR TB before his disease progressed to XDR TB during follow-up treatment with second-line drugs.
Patient 1 was a 33-year-old man who was born in Côte d'Ivoire, moved to Burkina Faso in 2000, and received a diagnosis of smear-positive pulmonary TB in July 2003. He received an 8-month standard treatment regimen (including 6 months of continuation with isoniazid and ethambutol). Treatment was directly observed during the fi rst 2 months. In January 2004, his treatment was classifi ed as failed, and he immediately started a standard retreatment regimen. Because his sputum did not clear by month 5, the patient was registered as a patient with chronic TB. During 2004, he traveled throughout Mali and Côte d'Ivoire. Back in Burkina Faso in January 2005, he was admitted to the reference national hospital for patients with chronic TB. His HIV test result was negative. At this time neither culture and DST nor a standard second-line drug regimen approved by the Green Light Committee were available in Burkina Faso. The patient was empirically prescribed kanamycin, ethionamide, ciprofl oxacin, and pyrazinamide, which he took under direct observation as a hospital inpatient for 21 months; at discharge, his sputum samples remained positive.
In December 2006, 2 sputum samples showed resistance to all fi rst-line drugs; resistance to second-line drugs amikacin, ofl oxacin, ethionamide, and cycloserine; and susceptibility to para-aminosalicylic acid and capreomycin. Appropriate and effective drugs were unavailable in the country. The patient was readmitted to the ward for patients with chronic TB and placed in a single isolation room. amikacin, kanamicin, ethionamide, and closerine but sensitive to ethambutol and ofl oxacin. In September 2007, the patient started treatment with Green Light Committee-approved second-line drugs (6 months of pyrazinamide, ofl oxacine, kanamycin, ethionamide, and cycloserine followed by 15 months of ofl oxacin, ethionamide, and cycloserine). Treatment was directly observed during the fi rst 13 months. His sputum samples remained positive throughout treatment, and additional DST in October 2008 showed an XDR strain of M. tuberculosis. Mycobacterial interspersed repetitive unit genotyping showed an identical pattern for strains detected initially and during follow-up.

Conclusions
Our fi ndings confi rm that XDR TB can be found wherever the search capacity exists. Thus, despite unavailability of evidence for the widespread existence of drug resistance, high priority should be given to strengthening laboratory capacities in sub-Saharan Africa (6). Also, because XDR TB developed while patient 2 was receiving second-line TB drugs, optimal adherence during intensive and continuation phases of second-line treatment regimens should be ensured. Staff should receive specifi c training with regard to managing the frequent side effects.
Our study had 1 major limitation. Because our sample was a select population and included patients already receiving treatment with second-line drugs, we may have overestimated the proportion of XDR TB cases among MDR TB cases.

Mechanisms of emergence of XDR TB in Burkina
Faso differ from those in South Africa, where most identifi ed XDR TB cases are primarily resistant, occur among HIV-infected patients, and result from exposure in the hospital or the community (7). Because each of the 2 XDR TB patients in Burkina Faso reported long-term stays in neighboring countries, we believe that response to the MDR TB challenge should be based on regional rather than national strategies.
Our study supports current policies for strictly controlled introduction of second-line drugs and the current Green Light Committee strategy that requires demonstration of compliance with guidelines for proper management of MDR TB before granting access to second-line drugs (8). Stakeholders and TB program managers in Africa should be reminded that the main reasons for development of resistance in TB are poor patient adherence to treatment regimen, inappropriate drug prescription, irregular drug supply, and poor drug quality (9). When these factors occur with use of second-line TB drugs, the result will be XDR TB.
Rapid adoption of a programmatic approach to management of MDR TB (6) is warranted in Burkina Faso and probably other countries in western Africa. In countries without effective programs for community or outpatient care of patients with MDR TB, hospital care in reference centers enables appropriate follow-up during the initial phases of treatment. In this context, implementation of appropriate infection control measures should rank high among program priorities (10).