9-month MDR-TB regimen cures 82% in African study

A 9-month standardised treatment regimen for multi-drug resistant tuberculosis (MDR-TB) cured 82% of people recruited to an observational study in Francophone Africa, the 46th Union World Conference on Lung Health heard on Saturday in Cape Town.

The regimen showed similar effectiveness to the performance of 9-month MDR-TB treatment regimens in previous studies in Africa and Bangladesh, and with the exception of hearing loss, serious adverse events were rare.

The finding gives further encouragement to the prospect of shortening MDR-TB treatment from 24 months to 9 months. A randomised study comparing 9-month and 24-month treatment for MDR-TB, the STREAM study, is due to report results in 2017. Further arms of the STREAM trial will test two additional MDR-TB treatment regimens: a 9-month all-oral regimen that does not require injections, and an even shorter 6-month regimen, both using the oral TB drug bedaquline.

The observational study was co-ordinated by the International Union Against Tuberculosis and Lung Disease and conducted in Benin, Burkina Faso, Burundi, Cameroon, Central African Republic, Côte d'Ivoire, Democratic Republic of the Congo, Niger and Rwanda. It recruited patients with confirmed rifampicin resistance, but excluded pregnant women and people who had received a previous second-line TB treatment regimen.

Participants received daily directly-observed therapy with:

• Four months of kanamycin, moxifloxacin, prothionamide, isoniazid, clofazamine, ethambutol, and pyrazinamide (normally abbreviated as 4Km Mfx Pto H Cfz E Z), followed by:
• five months of moxifloxacin, clofazamine, ethambutol and pyrazinamide (5 Mfx Cfz E Z).

Drug susceptibility was established by molecular or phenotypic testing and checked by supranational laboratories, at baseline and on isolates obtained at month 6 of treatment or later. Outcomes of treatment were assessed by smear and culture testing and by clinical examination each month during treatment and every six months after the completion of treatment.

The study recruited 1029 participants over 29 months; Christopher Kuaban presented preliminary results on 408 participants who began treatment prior to July 2014. 28% had experienced failure of standard first-line treatment with 2HREZ/4HR, 26% had failed the `category II` 8-month re-treatment regimen of 2HREZS/1HREZ/5HRE, 27% had relapsed after treatment and 14% were new cases of drug-resistant infection (5% were unclassified).

Just over half had major lung disease as a result of MDR-TB (covering > 50% of the lung area).

Smear and culture conversion were rapid for most participants. At month 3, 89% were culture negative and 77% were smear negative, and overall, intent-to-treat analysis showed that 82% were cured (defined as treatment termination without relapse and three consecutive negative cultures after a previous positive culture). The treatment failure rate was low (2.9%) but 6.6% were lost to follow-up and 7.8% died.

Overall, 22% of participants were HIV-positive and although treatment success rates did not differ by HIV status among those who survived, the death rate was higher in participants with HIV: 18% died, compared to 5% of HIV-negative participants.

High-level resistance to a fluoroquinolone was associated with a significantly lower cure rate (37.5% vs 83.7% for low-level resistance, p < 0.001), while high-level resistance to isoniazid was associated with a modestly reduced likelihood of cure when compared to low-level resistance (76.1% vs 87.8%, p > 0.05). Lack of susceptibility to fluoroquinolones was rare in this study population, affecting 8 participants, whereas high-level resistance to isoniazid was common, affecting just over a quarter of participants (n = 109).

The most frequent severe and serious adverse event was hearing loss, affecting 8% of participants. Grade 1 liver enzyme elevations and gastrointestinal side effects were common, as were kidney toxicity, but these adverse events were rarely reported as severe. Hearing loss was significantly more common in people who already had some evidence of hearing loss at baseline (23% vs 4%, p < 0.01) and in people living with HIV (17% vs 7%, p < 0.01).

“These preliminary results from using a 9-month MDR-TB treatment regimen are excellent,” said Arnaud Trebucq of The Union, a lead investigator of the study. “Implementing the shortened regimen is proving feasible and with improved outcomes compared with the standard MDR-TB treatment regimen.”