4.3 Considerations for implementation

The regimens recommended for treatment of Hr-TB do not have an intensive and a continuation phase – this simplifies the delivery and monitoring of treatment. Treatment is given daily, and intermittent treatment should be avoided. Relevant measures to support adherence, such as directly observed treatment (DOT), social support and the use of digital technologies should be considered to ensure favourable treatment outcomes (2).

The cost of medicines to compose a full 6(H)REZ regimen with levofloxacin is higher than the cost of a 2HREZ/4HR regimen used for drug-susceptible TB (25). Nonetheless, the 6(H)REZ regimen is an affordable and feasible intervention, even in low-income settings. Use of FDCs simplifies treatment and lowers costs, and the use of dispersible formulations of HRZ, ethambutol and levofloxacin is preferred in children. As with the treatment of other forms of TB, the expenses associated with the proper delivery of care (e.g. DST, adherence support and clinical monitoring) far exceed the cost of medicines.

There is currently no diagnostic platform approved for the detection of Hr-TB that matches the rapidity and convenience of Xpert MTB/RIF for rifampicin resistance, although a new Xpert cartridge that can detect isoniazid resistance is in the pipeline. First-line LPA can detect isoniazid resistance; it requires infrastructure that is typically available in a provincial or central level facility. Typical processing time for an LPA specimen is about 2–3 days, owing to batching. DST based on liquid culture (or MGIT) could also detect Hr-TB at the level of a reference laboratory, but this means an obligatory processing delay of at least 10 days. Testing on solid media is also an option, but it takes several months to obtain results; hence, this approach is of limited use for baseline testing and monitoring of treatment response.

Current epidemiological data indicate that more than three quarters of the global burden of Hr-TB occurs among previously untreated (“new”) TB cases. Previous TB treatment is thus not a strong indicator of risk of Hr-TB – the correlation with previous TB treatment is weaker than it is with MDR-TB. Reserving isoniazid DST to such patients is therefore unlikely to yield many Hr-TB cases. There are various concerns about empirical Hr-TB treatment of previously treated TB cases, without prior drug susceptibility testing. First, such treatment will lead to unnecessary overtreatment with fluoroquinolones and prolongation of pyrazinamide use in many patients. Most re-treatment cases will not have Hr-TB and can be cured with a 2HRZE/4HR regimen. Second, unless rifampicin resistance is excluded at the baseline, patients with MDR/RR-TB would be exposed to an inadequate regimen, with the risk of acquiring additional resistance, including fluoroquinolones. Third, this policy would deflect the focus of the programme from testing new (previously untreated) TB patients, who usually harbour the main burden of Hr-TB. Finally, this approach would risk creating once again a “re-treatment regimen”, similar to the situation that prevailed in many settings until recently with the indiscriminate use of the streptomycin-containing 8-month “Category 2” regimen in all previously treated TB patients.

In a situation where access to DST is good, a logical diagnostic algorithm would start with Xpert MTB/ RIF as the initial test for all patients evaluated for TB. Cases in whom TB is confirmed and rifampicin resistance is not detected would be further tested with LPA. Liquid culture may replace LPA, but the additional delay in getting results is a disadvantage.

Book navigation