Testing of resistance to isoniazid, pyrazinamide and fluoroquinolones is of paramount importance to guide treatment decisions and ensure that patients receive the best standard of care and have the highest chances to be cured. In this report the authors present an overview of the prevalence of resistance to isoniazid, pyrazinamide and fluoroquinolones and discuss the need of rapid diagnostics to detect resistance to these drugs
by Matteo Zignol
and Mario Raviglione
Global Tuberculosis Programme, World Health Organization
Resistance to Isoniazid, Pyrazinamide and Fluoroquinolones in Patients with Tuberculosis
Mycobacterium tuberculosis is today the top infectious killer with 1.8 million deaths caused by tuberculosis every year, including 390,000 among people with HIV infection. Every year, 10.4 million people are affected by tuberculosis. Furthermore, some 480,000 multidrug-resistant tuberculosis (MDR-TB)  cases emerge annually and an additional 100,000 cases of rifampicin-resistant tuberculosis, requiring the same treatment as MDR-TB, are estimated to arise in the same period of time. Drug-resistant tuberculosis represents a major threat to global health and may compromise the fight against tuberculosis in many countries.
Since 1994 the World Health Organization hosts a global project on surveillance of drug resistance in tuberculosis, which is the oldest and largest antimicrobial resistance surveillance project worldwide.
Very often when discussing of drug resistance in tuberculosis reference is made exclusively to MDR-TB and rifampicin-resistant tuberculosis, which are forms of tuberculosis requiring longer, more toxic and expensive treatment regimens with second-line drugs. However drug-resistance in tuberculosis is not exclusively MDR-TB or resistance to rifampicin.
Like rifampicin, isoniazid and pyrazinamide are the other two powerful drugs that are the pillars of the first-line regimen to treat tuberculosis. Fluoroquinolones (levofloxacin, moxifloxacin and gatifloxacin) represent the cornerstone of the second-line therapy to treat MDR-TB.
In this report we present an overview of the prevalence of resistance to isoniazid, pyrazinamide and fluoroquinolones and discuss the need of rapid diagnostics to detect resistance to these drugs.
Since current laboratory methods are sufficiently accurate for both, isoniazid resistance is usually investigated in parallel with rifampicin resistance: simultaneous resistance to both drugs defines a case of MDR-TB. Data from 155 countries for the period 1995−2015 collected either through continuous surveillance systems based on routine testing of all patients with tuberculosis or periodic surveys, which are discrete studies measuring drug resistance among a selected sample of patients who are representative of an entire population of tuberculosis patients, were used to estimate the prevalence of resistance to isoniazid without concurrent rifampicin resistance. Drug susceptibility testing results were available for 185,745 patients. The proportions of tuberculosis patients with strains that were resistant to isoniazid but susceptible to rifampicin were weighted by the number of new cases of tuberculosis that were notified in the country to generate regional and global averages.
Among all tuberculosis cases, the global average of isoniazid resistance without concurrent rifampicin resistance was 10.2% (95%CI: 8.4%–11.7%), ranging from 6.1% in the African region to 13.3% in the European region. With the expansion of the use of Xpert MTB/RIF testing (Cepheid, Sunnyvale, CA, USA) in many countries it is important to note that on average around one in 10 TB patients diagnosed to be sensitive to rifampicin will have isoniazid resistance that cannot be detected by the current Xpert MTB/RIF assay. These patients are likely to have poorer treatment outcomes and higher risk of acquiring rifampicin resistance during treatment compared to patients with fully susceptible tuberculosis, as shown in a recent systematic review of 3,744 patients with isoniazid-resistant tuberculosis (Gegia M et al. Lancet Infect Dis. 2017;17(2):223–234).
Pyrazinamide resistance is not routinely investigated as part of diagnostic and surveillance efforts in most settings because of the limitation of the currently available phenotypic test (MGIT 960) which is expensive, difficult to perform and poorly reproducible. For this reason, little information about the extent of resistance to pyrazinamide at the population level is available. Sequencing of the pncA gene of M. tuberculosis that expresses resistance to the drug is becoming an alternative method to investigate resistance to pyrazinamide. Data were gathered from surveys conducted in five countries (Azerbaijan, Bangladesh, Belarus, Pakistan, and South Africa) on a total of 4,972 patients. Levels of resistance varied substantially in the surveyed settings (3.0–42.1%) and resistance to pyrazinamide was always significantly associated to rifampicin resistance. Despite this finding, for a substantial proportion of patients with rifampicin-resistant tuberculosis (19–63%) pyrazinamide is still effective, confirming that this drug could still be used as part of the second-line regimen for the treatment of MDR-TB, as currently recommended by the World Health Organization.
Later generation fluoroquinolones, including levofloxacin and more importantly moxifloxacin and gatifloxacin (which are the newest molecules belonging to the 4th generation), are considered the most important drug class for the treatment of MDR-TB. Fluoroquinolones are currently being investigated in clinical trials to assess the efficacy of short rifampicin-sparing regimens for the treatment of tuberculosis (e.g. the NC-005 trial of TB Alliance). The prevalence of resistance to fluoroquinolones is unknown in many countries due to the limited laboratory capacity to perform second-line drug susceptibility testing. Data were gathered from surveys conducted in five countries (Azerbaijan, Bangladesh, Belarus, Pakistan, and South Africa) on a total of 5,015 patients. Overall levels of resistance to fluoroquinolones among all tuberculosis patients ranged from 0.5 to 12.4% for levofloxacin, and from 0.9 to 14.6% for moxifloxacin. Among patients with rifampicin resistance the prevalence of resistance to fluoroquinolones was significantly higher, ranging from 8.4 to 26.8% for moxifloxacin. Cross resistance between levofloxacin and moxifloxacin (when tested at 0.5 μg/mL) was nearly complete. High levels of fluoroquinolones resistance were detected in Pakistan, which can be the expression of extensive and unregulated use of fluoroquinolones in that country.
Testing of resistance to isoniazid, pyrazinamide and fluoroquinolones is of paramount importance to guide treatment decisions and ensure that patients receive the best standard of care and have the highest chances to be cured. Over the past few years progress has been made on the development of rapid molecular tests to detect drug resistance in tuberculosis, with Xpert MTB/RIF being the best example, given the limited technical requirements, relatively low cost and rapidity in producing results. However, at the moment, easy-to-perform, automated and rapid molecular tests to detect resistance are currently available only for rifampicin. The understanding of the genetic basis of drug resistance and of the correlation between genotypic and phenotypic testing results is rapidly improving. This should lead to the development of a larger range of molecular diagnostic tools capable to effectively detect resistance to several drugs, in particular to isoniazid, pyrazinamide and fluoroquinolones, in addition to rifampicin. Candidates exist and are undergoing testing in trials, although it may take a few more years before they become widely available. These rapid drug susceptibility tests are today a must as medical practice keeps evolving towards more precision and personalization. Without new rapid tests a large number of patients will continue receiving therapies which are often inadequate to combat the bacilli they harbor, with a higher risk of treatment failure, creation of additional resistance and continued transmission of the disease.
 MDR-TB is defined as a form of tuberculosis that is resistant to at least isoniazid and rifampicin