People living with HIV, those who are underweight and those
with more extensive drug resistance are more likely to die during treatment for
multi-drug resistant tuberculosis (MDR-TB), a nine-country study has found. The
findings were presented at the 46th Union World Conference on Lung
Health in Cape Town last week.
People diagnosed with multi-drug resistant TB are known to
have poor outcomes compared to people diagnosed with drug-sensitive TB, but
there are few data to provide a reliable assessment of the death rate in MDR-TB
patients, nor do clinicians have a detailed picture of which patients are at
greater risk of death and therefore in need of especially close monitoring
during MDR-TB treatment.
The Global Preserving Effective TB Treatment
Study (PETTS) was an international
observational cohort study designed to establish risk factors for
acquisition of TB drug resistance and the impact of acquired resistance on
treatment outcomes. Primary
results of the study were published in 2012. The analysis of the PETTS
cohort presented at the 46th Union World Conference looked at the
two-year mortality of people starting second-line TB treatment between January
2005 and December 2008 in nine countries in Africa, Asia, Eastern Europe and
Latin America.
The analysis included all adults recruited to the cohort
study who had received at least one month of treatment, and excluded pregnant
women, prisoners and children aged under 18, as well as anyone with
extrapulmonary TB.
The study evaluated participants according to five
pre-defined World Health Organization (WHO) criteria: cure, completed treatment, lost to follow-up,
treatment failure or death.
The study enrolled 1659 participants, of which 52 were
excluded from the final analysis due to missing outcomes data and 55 were
excluded due to lack of information on the time to outcome. The final analysis
comprised 1550 patients. The analysis censored the follow-up period after
achieving any outcome apart from death, and also censored participants who had
achieved no final outcome of treatment after two years.
Study participants had a median age of 37, 36% were women
and HIV prevalence was 18.5% – but in South Africa just over half of
participants was living with HIV.
Fifty-seven per cent of study participants were enrolled in
two countries – South Africa and the Philippines – but 62% of all deaths in the
study occurred in South Africa alone.
Overall, 191 people died during the first two years after
beginning treatment (14%). 13.9%
of patients who previously received first-line medications for TB died during
MDR-TB treatment; 11.5% of those who had received second-line medications
died. The death rate was especially high
among people with HIV (27.7%) and those who were underweight at baseline (BMI
< 18.5) (19.6%).
Those with
resistance to at least one fluoroquinolone and to at least one injectable
second-line drug (classified as extensively drug resistant, or XDR-TB) died more frequently than those with resistance to a
fluoroquinolone alone (28.6% vs 22.1%), but multivariable analysis found that
resistance to either a fluoroquinolone or to an injectable was independently
predictive of an increased risk of death during MDR-TB treatment (adjusted hazard
ratio 2.2, p < 0.001).
People who
were underweight (aHR 2.7, p < 0.001) or HIV-positive (aHR 1.8, p = 0.01) were
also at increased risk of dying during the first two years after beginning
treatment.
Country
differences in outcomes or risk factors were not presented. In particular the study did not report whether participants with HIV were receiving antiretroviral treatment at baseline, so it is unclear how this might affect outcomes.
Presenting
the findings, Hannah Kirking of the US Centers for Disease Control Division of
HIV/AIDS and TB said that clinicians should be aware of the factors amenable to
interventions that affect outcomes of MDR-TB treatment. Antiretroviral treatment
and interventions to increase body weight could have an impact on outcomes, as
could greater use of drug susceptibility testing to optimise MDR-TB treatment.