In May 2016, WHO recommended the use of a shorter multidrug-resistant TB (MDR-TB) regimen under programmatic conditions. The 9–12 months regimen consists of treatment with kanamycin, moxifloxacin, prothionamide, clofazimine, high-dose isoniazid, pyrazinamide and ethambutol followed by moxifloxacin, clofazimine, pyrazinamide and ethambutol. It is recommended for patients without previous exposure to second-line drugs (SLDs) and TB strains susceptible to fluoroquinolones and second-line injectable agents. It is not recommended for pregnant women and patients with extrapulmonary TB.1 This recommendation is based on observational studies2–4 as well as on the STREAM (Evaluation of a Standardised Treatment Regimen of Anti-tuberculosis Drugs for Patients with Multidrug-resistant Tuberculosis) trial performed in Asia and Africa.5

WHO estimates that in 2015 about 480 000 new MDR-TB cases occurred worldwide with 73 000 cases registered in the WHO European Region, where MDR-TB prevalence rates in new cases range between 10% and 34%.6 However, there were no studies on the potential use of the shorter regimen in the European Region to inform the guidelines. A recent retrospective investigation of a cohort of 1140 patients with MDR-TB in Western Europe showed that only 7.8% of patients with MDR-TB would have been eligible for this regimen.7

We aimed to assess the proportion of patients with MDR-TB that would be eligible for the shorter regimen in the Eastern European settings, where the rate of drug resistance is highest.

We revisited the data from 737 adult pulmonary MDR-TB cases in our cohort recruited in 2007 and 2009 and followed-up to 2012 across Latvia, Lithuania, Estonia and Bucharest city within the European Union-funded project ‘TB-PAN-NET’.8 Drug susceptibility testing (DST) for first-line drugs (FLD) and SLD was performed using solid or automated liquid culture media systems (BACTEC MGIT 960, Becton Dickinson, Sparks, Maryland, USA) according to standard procedures.9 High and low levels of isoniazid resistance were not distinguished phenotypically.

Fluoroquinolone resistance was defined based on resistance to ofloxacin or moxifloxacin; DST to ethionamide or prothionamide was equivalent. Clofazimine DST was not routinely conducted. Eligibility for the shorter regimen was defined as susceptibility to kanamycin, moxifloxacin/ofloxacin, ethionamide/prothionamide, pyrazinamide and ethambutol with no history of previous treatment. Non-eligibility was defined when at least one of the exclusion criteria was valid irrespective of the overall information completeness per case. Descriptive statistical analyses stratified by eligibility were performed using Stata V.14 (StataCorp, College Station, Texas, USA).

Among 737 patients with MDR-TB, 610 (83%) had sufficient DST information for the evaluation. Among these, 38 (6.2%) were eligible for the shorter regimen and 572 (94%) were non-eligible. Drug resistance to ethambutol (64%) or pyrazinamide (58%) was the most common reason for exclusion; however, concerns regarding DST reliability might question the findings. Considering both DST and information on previous treatment, 687 (93%) had sufficient information for the evaluation. Among these, only 29 (4.2%) were eligible and 658 (96%) non-eligible (table 1). The eligibility was largely similar across different patient groups. However, none of the patients with TB living with HIV or using recreational drugs was eligible (table 2).

These results are largely in line with published data for other regions; the proportion of 6.2% eligible patients was slightly higher than in pooled findings from the European and Latin American settings (4.0%)10 and slightly lower than in a Western European cohort (7.8%).7 Of note, even in the context of settings with a well-developed quality controlled DST, many DST profiles were not complete.

The study was limited to the information available for the cohort that was based on routine management data. Complete information on the level of isoniazid resistance, pregnancy and drug intolerance was missing. However, over 90% of all isoniazid-resistant strains (mainly of the Beijing family) isolated in many Eastern European countries are known to harbour mutations in katG gene associated with high levels of isoniazid resistance.11 ,12

Previous studies have demonstrated that the epidemic in the former Union of Soviet Socialist Republics countries including the Baltic countries is likely to be associated with a fairly homogenous ‘East European’ subtype of the Beijing lineage.13 We therefore hypothesise that the findings might be applicable to the countries beyond the study settings.

In conclusion, in many Eastern European countries, only an extremely small proportion of patients with MDR-TB would be eligible for the shorter regimen. Given complex drug-resistance patterns and the expected significant prevalence of high-level resistance to isoniazid, the shorter regimen should be applied with great caution. It is crucial to consider all exclusion criteria before its administration. Reliable and comprehensive DST must be available before considering the new regimen for programmatic use. Meanwhile, available molecular assays, specifically those detecting mutations in genes associated with isoniazid, fluoroquinolones and aminoglycoside resistance can support the adequate design of drug regimens. Further larger research in different geographic settings is critically needed to assess benefits of the shorter regimen. The impact of the new regimen on adherence improvement should be evaluated. Should the shorter regimen be proved to be applicable at least for a proportion of patients, costs saved due to the shortened treatment duration could be invested in measures targeted to improve adherence.

Overall, in spite of the demonstrated success of the shorter MDR-TB regimen in Africa and Asia, its use in Eastern Europe might be an exception rather than a rule. The continuous availability of all required drugs in the regimen needs to be addressed in settings with limited resources. Further options for shorter regimens are still urgently needed for most patients with MDR-TB.