Abstract
Collaboration is needed to encourage and optimise use of effective drugs and their impact on overall DR-TB treatment http://ow.ly/TRkcu
From the authors:
We thank Bhuniya and colleagues for highlighting these common concerns regarding the use of linezolid in drug-resistant tuberculosis (DR-TB). We confirm that patients in our cohort were initially started on treatment with standardised regimens following diagnosis of rifampicin and isoniazid resistance (multidrug-resistant tuberculosis (MDR-TB)). We did have access to drug susceptibility testing (DST) for first- and second-line drugs (rifampicin, isoniazid, ethionamide, fluoroquinolones and injectable agents) and regimens were only individualised following detection of second-line drug (SLD) resistance or failure of standardised regimens.
Individualised regimens were designed based on prior history of antituberculosis drug use and results of first- and second-line DST, as endorsed by World Health Organization (WHO) guidelines; all available drug options, including clofazimine, high-dose isoniazid, amoxicillin/clavulanate and bedaquiline, were used where possible. We emphasise that for all patients in this cohort, the recommended standardised MDR-TB regimens in either country were not considered adequate to provide treatment with at least four likely effective drugs, which is one of the WHO's basic principles of treatment [1].
While we agree with Bhuniya and colleagues that it is important to preserve the efficacy of newer and repurposed antituberculosis drugs such as linezolid (LZD), we also advocate greater access to these drugs in regimens containing at least two other likely effective drugs, within optimally administered treatment programmes with sufficient monitoring and adherence support [2]. Treatment outcomes are poor for MDR-TB, with <50% treatment success globally and in both settings included in our study [3, 4]; outcomes are even worse for patients infected with strains with additional second-line resistance [5]. Persistent use of standardised regimens that are potentially inadequate in cases with known SLD resistance, or where standardised treatment has already failed, is likely to contribute to the emerging extensively drug-resistant tuberculosis (XDR-TB) epidemic due to further acquisition of SLD resistance through selective drug pressure [6], particularly in high-burden settings. This defines the need for strengthened treatment regimens including new TB drugs, both to improve individual patient outcomes and to prevent undertreatment on a programmatic level that drives further resistance.
As ours was a small, retrospective study among DR-TB patients offered treatment with LZD in high-burden programmatic settings, and not a randomised controlled trial to prove the efficacy of LZD; there was no clear control group to allow comparison of our findings. Given their poor prognosis, we optimised treatment using any available medications thought to be effective, in order to offer patients a reasonable chance of cure. There have been very few reports of LZD use for tuberculosis/HIV-coinfected patients; our interim outcomes appear more favourable than previous reports on XDR-TB treatment outcomes for predominantly HIV-uninfected patients treated with standardised regimens [7].
Bhuniya and colleagues suggest that use of LZD at improper dosages and duration is associated with acquired resistance. The article they refer to in support of their argument in fact reports a high proportion of LZD resistance among isolates taken prior to any LZD treatment and emphasises the need to assess resistance early [8]; this is not a reason to avoid LZD in DR-TB treatment. While 11% of patients in a recent randomised controlled trial of LZD in XDR-TB developed linezolid resistance, those patients were effectively receiving monotherapy, unlike our cohort [9]. Furthermore, there is in vitro evidence that LZD resistance occurs at lower frequency than what is observed with other tuberculosis drugs [10]. Drug exposure of LZD is highly variable, as is drug susceptibility, with wild-type minimum inhibitory concentration ranging from <0.125 to 0.5 mg·L−1. While there is no existing evidence to suggest that therapeutic drug monitoring (TDM) for LZD improves treatment outcomes, it is possible that TDM-guided dosing may prevent toxicity and development of acquired drug resistance. Although TDM of tuberculosis drugs may be challenging in resource-limited programmatic settings, this could be facilitated by innovations such as dried blood spot testing [11, 12].
LZD is still classified as a group 5 drug by WHO and is often only recommended for XDR-TB; however, there is substantially more evidence supporting the effectiveness of LZD than currently exists for many other SLDs used routinely in standard DR-TB treatment regimens [9]. Despite the frequency of adverse events associated with long-term use of LZD, these can be appropriately monitored, and we support wider use of this drug precisely due to the lack of “other, safer alternative” treatment options that would still be effective. The risk of toxicity and the demonstrated clinical effectiveness of LZD should be carefully weighed against the adverse events and monitoring requirements associated with other, less effective SLDs. Joint collaboration between tuberculosis practitioners is necessary to encourage and optimise appropriate use of effective drugs and their potential impact on overall DR-TB treatment outcomes, rather than severely restricting their use and limiting their potential wider public health benefit.
Footnotes
Conflict of interest: None declared.
- Received August 18, 2015.
- Accepted October 24, 2015.
- Copyright ©ERS 2015