Shorter regimens for multidrug-resistant tuberculosis should also be applicable in Europe

To the Editor:

Since May 2016, the World Health Organization (WHO) has recommended shorter regimens to treat rifampicin-resistant/multidrug-resistant tuberculosis (RR/MDR-TB), substantially reducing the treatment duration to 9–12 months [1]. van der Werf et al. [2] estimated that only 11% of RR/MDR-TB patients in the European Union (EU)/European Economic Area (EEA) fulfilled the WHO inclusion criteria, which is similar to estimates by Lange et al. [3] and Sotgiu et al. [4]. This estimate raises concerns, as that the conventional long-duration regimen has very poor results in the EU/EEA. Health sector crisis, increasing inequality and xenophobia add to the urgency to prevent and manage MDR-TB. The main exclusion criteria are as follows.

• 1) Previous use of second-line drugs (SLDs). The estimate is clearly too high (55% of RR/MDR-TB cases), as it also excluded previous use of only first-line drugs. A better estimate could be from countries like Latvia where 42% of RR/MDR-TB cases in 2005–2015 were previously treated, 30% of them with SLDs, representing 12% of all RR/MDR-TB cases (V. Riekstina, Centre of Tuberculosis and Lung Diseases, Riga, Latvia; personal communication). In Western Europe, many patients with RR/MDR-TB were born in Asia and Africa, where SLDs have been used far less frequently [5].

• 2) Resistance to second-line injectable drugs (SLIDs) and/or fluoroquinolones. Such resistance affects 41% of RR/MDR-TB patients in the EU/EEA; 34% in new patients and 48% in previously treated patients (communication with the European Centre for Disease Prevention and Control (ECDC), Stockholm, Sweden) (table 1). In patients of Former Soviet Union (FSU) origin, resistance levels are higher (50% in all, 44% in new and 56% in previously treated patients). In patients whose origin was outside the FSU (64% of all cases), the level of resistance was considerably lower (34% overall, 22% in new and 42% in previously treated patients). The proportion in previously treated patients is even lower, if patients treated with SLDs are excluded first (criterion 1). Furthermore, all ofloxacin-resistant cases were also considered to be moxifloxacin-resistant, but only 7% of ofloxacin-resistant strains have been found to be resistant to moxifloxacin [6]. In a study performed in Bangladesh, only high-level fluoroquinolone resistance reduced success, caused by failure [7]. Most of the mutations correlating with this high-level resistance can be identified by the line-probe assay recommended by the WHO [8].

• 3) Resistance to other drugs in the regimen, particularly ethambutol or pyrazinamide [1]. In the cohort studies performed in Bangladesh [7] and West Africa [9], on which the recommendation for the shorter regimens is based, cases with resistance to isoniazid, ethambutol or pyrazinamide were not excluded and the overall success was high. Moreover, WHO does not recommend that treatment decisions should be based on drug-susceptibility test (DST) results for ethambutol as the test is unreliable [10]. There is no approved rapid test for pyrazinamide resistance and the clinician may decide to use the shorter MDR-TB regimen in its presence [10]. Susceptibility to other drugs, such as clofazimine, is either never tested routinely, or their DST results are notoriously unreliable and their testing is therefore discouraged. In the EU/EEA, DST coverage is fortunately high (fulfilling WHO criteria for exclusion of resistance to SLID and fluoroquinolones), meaning resistance to ethambutol and pyrazinamide would therefore be a major reason for exclusion.

• 4) An extrapulmonary site, with exclusion based on lack of data. The rifampicin-based short-course chemotherapy was first developed for pulmonary TB in the 1970s and 1980s [11], and later studies found that it was also effective in extrapulmonary forms (with the exception of TB meningitis) [12]. There is no apparent reason to postulate that the shorter MDR-TB regimen will not work in extrapulmonary TB.

Shortened MDR-TB regimens have been shown to protect against acquired resistance to SLD (one in 515 cases in Bangladesh) [7]. As they do not include new drugs, they will be protected and provide life-saving treatments for patients with resistance or adverse reactions to SLDs. By the end of 2015, bedaquiline had already been used in 70 countries; failures and resistance development has been reported [13].

Specialists may be reluctant to use standardised regimens [14]. When DOTS (directly observed treatment, short course) was introduced globally in the 1990s, standard regimens had already long been adopted in leading European countries without diminishing the role of specialists in providing holistic treatment for their patients. It could and should be analogous with the standardised shorter MDR-TB regimens. Specialists may treat most of the patients with shorter regimens at low cost, leaving more time and funding for the increasing number of patients with resistance to SLID or fluoroquinolones, as well as adverse reactions.

In Norway, the National MDR-TB Technical Group recently included the shorter regimens in the treatment recommendations (unpublished data). WHO criteria are generally interpreted to exclude most patients because of resistance to ethambutol, pyrazinamide or an extrapulmonary site. This forces the patient to be enrolled on the conventional regimen with twice the duration and documented unacceptable results. The shorter regimens should also be eligible for patients with strains resistant to ethambutol or pyrazinamide (not to contradict the latest WHO recommendation) and should be evaluated in patients with an extrapulmonary site (under operational research conditions) in EU/EEA countries, adding much needed evidence about the regimens from another setting.