Treatment of brucellosis

Although human brucellosis has been recognised for 121 years it remains difficult to treat.1 It is transmitted mainly from domestic animals to humans through direct contact, contaminated animal products (particularly dairy products), and by inhalation of infectious particles. Brucella has developed many ways to evade the human immune system, and it induces a disease that is often relapsing or chronic. The geographical distribution of the disease is constantly changing, with new foci emerging,2 and Brucellaalso has the potential to be used in biowarfare as it is easily produced in a steady aerosolised form.3 Brucella’s unique interaction with the human immune system means that a protracted therapeutic regimen with a combination of antibiotics is needed to avoid treatment failure and relapses, serious complications, or residual damage from focal disease.4 5 The optimal treatment regimen is debatable.

In the accompanying paper, Skalsky and colleagues report a systematic review of randomised controlled trials of different antibiotic regimens used for human brucellosis.6 Conducting such a review is no easy task because the trials assessed many combinations of antibiotics, which were given under different clinical conditions, for different periods, and for infections induced by strains with undefined and potentially different virulence.

Many of the trials recruited small numbers of patients, compared suboptimal regimens, used dubious diagnostic criteria, and inadequately evaluated side effects (for example the nephrotoxicity and ototoxicity of aminoglycosides). Moreover, trials of brucellosis treatment have inherent pitfalls related to how therapeutic success is defined and the need for a long follow-up.7 Skalsky and colleagues have done an excellent job of tackling these shortcomings.

The authors conclude that a triple regimen of doxycycline, aminoglycoside, and rifampicin is the optimal combination (relative risk of failure compared with doxycycline-aminoglycoside 0.40, 95% confidence interval 0.20 to 0.79). This conclusion is based on just two randomised controlled trials, the first of which was undertaken in patients with spondylitis—a complication of brucellosis that is notoriously difficult to treat.5 Most specialists agree that brucellar spondylitis has to be targeted aggressively and early to minimise harm, although this does not apply to the treatment of brucellosis in general. The second trial, in contrast, found no significant difference between double and triple regimens.

So how should we interpret these results? Eventually it goes back to the long standing dilemma about treating brucellosis—whether to combine doxycycline with a parenteral aminoglycoside or with rifampicin. Skalsky and colleagues’ conclusions emphasise previous observations that it is best to include an aminoglycoside in the therapeutic regimen (either streptomycin or gentamicin, although for gentamicin the duration of administration is still vaguely defined).

Despite the thorough analysis it is unclear how the results of the review translate into everyday clinical practice. A study of patients with brucellosis showed that—even when told about the shortcoming of rifampicin compared with aminoglycoside—most patients preferred to use a convenient all-oral combination, although the risk of relapse is higher.8 Furthermore, even when doctors are aware of the superiority of the regimen containing aminoglycoside, most prescribe the convenient but inferior regimen.9 Convenience may lead to better adherence, and thus improved overall outcome, which may be crucial for a disease that requires many weeks of treatment.

The review also finds that longer durations of treatment (six weeks) significantly reduce relapses and treatment failure compared with shorter durations (less than 30 days). Rates of relapse could possibly be reduced further if regimens were continued for longer than six weeks.

Cure in brucellosis is defined purely on clinical grounds by the absence of symptoms and signs of residual or relapsing disease. In addition, the definition of chronic disease and its parameters are vague. Recent studies, however, have shown a residual bacterial load in a considerable proportion of clinically healthy patients who have had brucellosis, even one to three years after “cure.”10 If such a pathogenic process exists, in future we may have to redefine what is meant by treatment success or cure.

The need for a global collaboration to tackle the medical, social, and political aspects of brucellosis has never been greater. The coalition of experts that was initiated in Ioannina, Greece, in 2006, which resulted in the “Ioannina recommendations,”11 and the publication of research such as that by Skalsky and colleagues are paving the way forward. Global clinical trials recruiting thousands of patients and the development of a global Brucella database are currently under way.