CROI: MDR TB cases in South Africa - person-to-person spread likely to be chief cause

Further evidence that infection control in South African hospitals is vital for containment of the country’s growing epidemic of drug-resistant tuberculosis was presented to the Fifteenth Conference on Retroviruses and Opportunistic Infections on Wednesday, with the finding that every case of multi-drug resistant TB analysed in one rural hospital was the result of re-infection, not poor drug adherence.

Multi-drug resistant TB is a growing problem in South Africa, and in 2006 researchers revealed that an outbreak of extensively drug-resistant TB had claimed over 50 lives in less than a year in one hospital in the province of KwaZulu-Natal.

Since that report drug sensitivity testing of TB bacteria isolates in the province have revealed that in 2006 alone, 2476 patients had multi-drug resistant (MDR) TB – compared with just 124 cases in the whole of the United States in 2006.

Five hundred and thirty-nine cases of MDR TB were reported between June 2005 and August 2007 from one hospital alone – the Church of Scotland hospital in Tugela Ferry, a small town deep in the rural uplands of KwaZulu-Natal.

The Church of Scotland hospital was the site which first identified the outbreak of extensively drug-resistant (XDR) TB, although not the source for an epidemic that became apparent throughout South Africa once TB specialists began a systematic search for extensive drug resistance in September 2006.

Prior to the Tugela Ferry outbreak, the conventional wisdom in the TB world had been that MDR TB was largely a product of poor adherence to the six-month regimen of TB drugs required to cure the disease. World Health Organisation advice has long emphasised the importance of directly observed therapy to ensure that TB drugs are taken.

However, research at the Church of Scotland hospital indicated that the XDR TB outbreak was attributable to airborne infection with the drug-resistant strain, often in the hospital.

The findings led to a call for better infection control in crowded hospitals that may house dozens of TB patients alongside hundreds of HIV-positive patients who are highly susceptible to rapid TB progression if they contract the infection. In Tugela Ferry the roll-out of antiretroviral therapy had paradoxically assisted the spread of XDR TB; one cluster of patients who died rapidly from XDR TB were members of an ARV support group, and it was the rapid decline of several patients previously doing well in this group that alerted health care workers that something was amiss.

But even after the warnings about the need for improved infection control, hospitals in southern Africa have been slow to change their ways.

In order to investigate the risk of MDR TB infection more systematically, Dr Neel Gandhi of Albert Einstein Hospital, New York, set out to look at patients with a prior history of TB who had eventually been diagnosed with MDR or XDR TB, to see what proportion were becoming superinfected with drug-resistant strains, as opposed to suffering the consequences of TB treatment failure.

Using retrospective medical records from the Church of Scotland hospital, Dr Gandhi and colleagues looked for patients with a prior TB diagnosis who had subsequent evidence of a second diagnosis of MDR or XDR TB, confirmed by drug sensitivity testing.

The researchers identified 17 patients, 88% of them HIV-positive, who had been identified with MDR or XDR TB a median of 150 days after a previous TB isolate had been found to be drug-sensitive.

The second isolate was available in each case because doctors had spotted signs of deterioration, often after several months of clinical improvement on TB treatment. Adherence to TB treatment was considered to be good in all cases, Dr Gandhi told reporters.

They carried out genetic matches of the drug-sensitive and drug-resistant TB strains to determine whether they had evolved, or whether the drug-resistant strain was completely unrelated. In every single case the drug-resistant strain was significantly different, and represented evidence of an `exogenous` infection – superinfection with drug-resistant TB.

Fifteen of the 17 patients died within 14 days of the collection of the sputum sample that subsequently yielded the drug-resistant TB isolate – invariably prior to the return of a diagnosis of MDR TB from the laboratory.

Speaking at a press conference, Dr Gandhi said the findings underlined the critical importance of infection control measures in containing MDR and XDR TB. He said that wards containing 40 to 70 beds were commonplace, and that hospitals needed to look at ways of identifying TB patients promptly and separating them from others, as well as ensuring good infection control in settings where TB patients were grouped together.

The patients who developed drug-resistant TB in this study had spent a median of 25 days in hospital, he pointed out.

He described how, at his hospital, a `cough officer` was responsible was identifying coughing patients on the wards, separating them from others, ensuring that they wore a surgical mask, and referring them to a doctor for urgent evaluation of their TB status.

Minimising hospital admissions and the time spent in hospital could also help, as could the introduction of outdoor waiting areas.