Mycobacterial disease

What the evidence shows

Drug-susceptible tuberculosis can now be treated in four months — but only with the right regimen. Rifapentine with moxifloxacin was non-inferior to six months of standard therapy (Study 31/A5349); rifapentine without moxifloxacin was not. An earlier attempt at four-month treatment using moxifloxacin alone cleared cultures faster and relapsed more (REMoxTB) — the clearest demonstration in the field that early bactericidal activity does not predict cure.

In children, four months matched six for drug-susceptible, non-severe, smear-negative disease, with 3% unfavourable outcomes in both arms (SHINE) — a population previously treated by extrapolation from adult trials.

Drug-resistant disease has been transformed. Bedaquiline, pretomanid and linezolid gave a 90% favourable outcome in extensively drug-resistant tuberculosis (Nix-TB), replacing 18 to 24 months of injectable therapy. Peripheral neuropathy affected 81%, so the dose was then optimised: 600 mg of linezolid for 26 weeks preserved 91% cure and cut neuropathy to 24% (ZeNix). Adding moxifloxacin for rifampicin-resistant disease produced an unfavourable outcome in 11% versus 48% with standard care (TB-PRACTECAL) — a 37 percentage point difference that says as much about how poor the old regimens were as about how good the new one is.

Non-tuberculous disease has one randomised treatment. Inhaled liposomal amikacin tripled culture conversion in refractory Mycobacterium avium complex infection, from 8.9% to 29.0% (CONVERT) — at the cost of respiratory adverse events in 87% of those treated, and on a microbiological rather than a symptomatic endpoint.