XDR TB emerging despite good adherence, due to sub-optimal treatment and bad infection control

Extensively drug-resistant tuberculosis is most likely emerging as a result of delays in diagnosis, sub-optimal treatment and poor infection control, not failures in patient adherence, a newly published study of XDR TB cases in a South African gold mine suggests.

Drug-resistant TB is an emerging concern in South Africa, and in 2006 a cluster of XDR TB cases was identified in KwaZulu Natal, leading to national attempts to implement improved infection control and drug susceptibility testing.

However, drug-resistant TB remains a major challenge for South African and other nations in the region, where high rates of tuberculosis are being exacerbated by HIV infection, which makes people exposed to TB much more likely to develop active disease.

The study, published this week in the US Centers for Disease Control online journal Emerging Infectious Diseases, looked at 128 patients diagnosed with drug-resistant TB at a single gold mine in the North West province of South Africa between January 2003 and November 2005.

Although the gold mine had achieved a cure rate above 85% in the treatment of new smear-positive cases by 2001, following the introduction of directly observed treatment that ensured optimal adherence, new cases continued to rise and the incidence of drug-resistant TB was also growing.

Investigators from Stellenbosch University, Harvard University School of Public Health and West Vaal Hospital identified 128 drug-resistant TB cases among 3003 TB cases notified at the gold mine’s hospital. Eighty-four were HIV-positive, seven were HIV-negative and the remainder had an unknown HIV status. Sixty-two per cent of those with MDR TB had a CD4 count below 200, indicating advanced HIV infection, but only 8% were already receiving antiretroviral therapy at the time of MDR TB diagnosis.

Only 31% of those who were diagnosed with MDR TB were cured; 35% died, with the remainder still receiving treatment at the time of the analysis.

The investigators found that 74 mTB isolates showed similarity to at least one other isolate, indicating transmission, and they were grouped in 11 transmission clusters, the largest of which involved 42 patients. The investigators calculated that 71% of patients had acquired drug-resistant TB from another patient.

Multi-drug resistant isolates were more likely to be found in clusters than single-drug resistant isolates; strains that were classified as XDR TB or pre-XDR (MDR with resistance to the second-line agents kanamycin or ofloxacin) were 27 times more likely to be found in a cluster than isolates with less drug resistance (OR 27.42, p<0.001), and no additional risk factors for clustering were identified in the multivariate analysis.

In the largest cluster, of 42 patients, three-quarters of patients had been hospitalised for non-MDR TB (in a general TB ward) at the same time as another patient in the cluster was admitted to the hospital with MDR TB care. Since all patients with TB received care on the same ward until diagnosed with MDR TB, it is not hard to see how such a large cluster could have emerged within a single facility.

In addition 92% of MDR TB patients in the cluster had worked in the same mine shaft as another person in the same MDR TB cluster, and 85% lived in the same dormitory as another MDR TB case prior to diagnosis.

Fifty-nine per cent of MDR TB patients in any cluster had a previous history of TB treatment.

The authors make a number of recommendations for curtailing the spread of MDR TB within hospitals and workplaces:

• Ensure that everyone who is eligible for antiretroviral therapy is getting it, in order to reduce the number of individuals susceptible to developing active TB.
• Greater efforts to identify infectious cases through intensified case finding, active screening and improved education about TB symptoms.
• More frequent sputum smear examinations to identify infectious cases, and more frequent culture-based diagnosis to identify cases before they become infectious.
• Development and implementation of rapid drug susceptibility testing in order to identify MDR cases, plan appropriate treatment and separate from susceptible patients.
• Better infection control measures on general wards, outpatient waiting areas and TB wards.

Full details of the methodology are available in the published paper, which is freely available at the journal website.