Abstract
Only limited evidence exists to support current recommendations for the treatment of human disease due to M. bovis http://ow.ly/h92K3023fLu
To the Editor:
Mycobacterium bovis, a member of the Mycobacterium tuberculosis complex, is an important cause of disease in cattle, but it can also cause disease in humans [1]. Transmission to humans generally occurs after close contact with infected animals or consumption of unpasteurised contaminated dairy products [1, 2]. The symptoms of human disease due to M. bovis are similar to those of disease caused by M. tuberculosis, although M. bovis is more likely to cause extrapulmonary disease [3]. In clinical practice, M. bovis can only be differentiated from M. tuberculosis using biochemical or genetic tests [3, 4].
Although human M. bovis disease is now rare in high-income countries, it is believed to be a more important health issue in low-income countries due to the lack of veterinary control measures and pasteurisation of milk [2, 3]. It has been estimated that M. bovis accounts for <1.5% of all human tuberculosis (TB) in regions outside Africa and ∼2.8% of all TB disease in Africa [5]. However, the incidence of M. bovis TB may be underestimated, because of the similarity of clinical features to TB caused by M. tuberculosis and because testing for M. bovis is not performed routinely.
As recommended by the United States Centers for Disease Control and Prevention, treatment of disease due to M. bovis usually consists of rifampicin, isoniazid and ethambutol [6]. Treatment duration is generally extended to 9 months due to the exclusion of pyrazinamide, since all strains of M. bovis are resistant to it. In the past two decades, while several reviews have investigated the epidemiology of M. bovis in humans [2, 5, 7], no review has evaluated the treatment regimens and outcomes of disease due to M. bovis. Therefore, we conducted a systematic review of all published articles that reported drug regimens and results of treatment of human disease due to M. bovis.
The search was conducted in three electronic databases: MEDLINE (through OVID), Embase (through OVID) and the Cochrane Library, from the start date of each database until the date of the search (September 8, 2015). Medical subject heading terms and keywords (in title and abstract) related to M. bovis, drugs of interest and treatment outcomes were used for the search. The following inclusion criteria were used: 1) there was no language restriction in the search, but the full-text had to be in English, French, Chinese, Portuguese or Spanish; 2) M. bovis disease that was confirmed by genetic tests or biochemical tests, but not disease caused by M. bovis bacille Calmette–Guerin; 3) cohort or randomised trial; 4) end-of-treatment outcomes, as defined by the World Health Organization (WHO), were reported [8].
Study characteristics and treatment outcome information were abstracted, including country, study period, study participant number, age, sex, HIV prevalence, extrapulmonary TB, drug resistance pattern, drug regimen, treatment duration and end-of-treatment outcomes. End-of-treatment outcomes were recorded according to WHO categories: success (cure and complete), failure, death, loss to follow-up and relapse. Authors of the studies included were contacted to obtain additional information.
Due to differences in regimens, results were not pooled. The treatment success rate of each cohort was calculated as: success/(success + fail + relapse) or success/(success + fail + death + relapse + loss to follow-up).
985 publications were identified through the initial database search, of which 17 were selected for full-text review. Of these, 14 were excluded. Six did not describe treatment regimens, two were conference abstracts and one was in Russian; five reported regimens and outcomes but were excluded for different reasons, as follows. Cicero et al. [9] reported 6-month status, rather than end-of-treatment outcomes; Kataria [10] and O’Donohue et al. [11] reported time to culture conversion; Esteban et al. [12] included multidrug-resistant M. bovis cases (seven out of 13) and used seven different regimens without reporting results by regimen; and Sauret et al. [13] reported treatment durations of isoniazid-rifampicin-ethambutol that varied from 4 to 12 months without stratifying results.
This left three studies which reported 439 patients with M. bovis disease from the USA, Argentina and the Netherlands, of whom 54.4% were male, and median ages were 42 years, 45 years and 62 years, respectively (table 1). In the two studies that reported HIV status, 39 out of 74 subjects and eight out of 37 subjects tested were HIV-positive. More than half of the patients (59%) had extrapulmonary disease, among whom lymph node disease was the most common form.
After excluding the patients who died, were lost or received other or unknown treatment regimens, two cohorts with 156 patients received the isoniazid-rifampicin regimen for 6–9 months and two cohorts with 113 patients received the isoniazid-rifampicin-ethambutol regimen for 6–12 months (table 1). When the denominator for success rate included all poor outcomes (fail + death + relapse + loss to follow-up, which might be considered equivalent to an “intention to treat” analysis), success was 74% for the isoniazid-rifampicin regimen and 79% for the isoniazid-rifampicin-ethambutol regimen. When the denominator included only success + fail + relapse (equivalent to efficacy), success rates were 99% and 93% for the isoniazid-rifampicin regimen and the isoniazid-rifampicin-ethambutol regimen, respectively. The reasons for the differences between these two methods of calculation of success rate were the high death rates (overall 15%) with rates of loss to follow-up contributing the remainder (6%). The impact on treatment outcomes of 6 months’ or 9 months’ duration of use, or from the added use of ethambutol could not be estimated due to the limited number of studies and patients. The major reason for low success rates when all outcomes were considered was the high mortality in all three studies. In two studies (LoBue and Moser [14] and Cordova et al. [15]), this high mortality could have been due to the high rates of HIV co-infection in the patients.
The current study has several limitations, the most important of which is that only three studies could be identified for this review, reflecting the rarity of this condition, at least in settings where testing for M. bovis is performed routinely. In addition, we could not pool results across studies, due to differences in regimen and duration, allowing only a simple comparison of rates. Fluoroquinolones were used in only one study (Majoor et al. [16]) and only for a small number of patients, preventing us from estimating results with these agents. The large proportion of patients treated for extrapulmonary disease meant that treatment completion, rather than cure, was measured in the majority of patients; this may have resulted in an overestimation of treatment success rate. An additional limitation was that some patients had underlying drug-resistant strains, which could have resulted in higher failure and mortality rates. HIV co-infection may also have contributed to the higher mortality rates, compounding the limitations in the interpretation of treatment outcomes with different treatment regimens.
In summary, there is very limited evidence supporting current recommendations for the treatment of human disease due to M. bovis. However, these limited data suggest that currently used regimens of isoniazid-rifampicin or isoniazid-rifampicin-ethambutol are adequate, although the benefit gained by adding ethambutol to isoniazid-rifampicin remains unclear. There were inadequate data to support a shorter duration of treatment of <9 months. Although better evidence to inform treatment recommendations for M. bovis would be welcome, a greater research priority would be to correct the paucity of accurate epidemiological and surveillance data in order to define the importance of M. bovis as a cause of human disease.
Acknowledgements
The authors thank Philip A. LoBue (Division of Tuberculosis Elimination, Centers for Disease Control and Prevention, Atlanta, GA, USA), Christof J. Majoor (Academic Medical Center, Amsterdam, the Netherlands) and Ezequiel Cordova (Infectious Diseases Unit, Hospital Cosme Argerich, Buenos Aires, Argentina) for providing additional information.
Footnotes
Support statement: This work was funded by a grant from the World Health Organization, which received funding from the United States Agency for International Development. Funding information for this article has been deposited with the Open Funder Registry.
Conflict of interest: Disclosures can be found alongside this article at erj.ersjournals.com
- Received March 29, 2016.
- Accepted June 23, 2016.
- Copyright ©ERS 2016