Melioidosis is a potentially fatal infection caused by the Gram negative bacillus, Burkholderia pseudomallei, following an encounter with contaminated soil or surface water. Like other environmental bacteria, B. pseudomallei is highly resistant to many antibiotics. The antibiotics normally used for first line treatment of Gram negative septicaemia (e.g., ampi-/amoxycillin, aminoglycosides, ceftriaxone) are ineffective in the treatment of melioidosis, highlighting the need for recognition of melioidosis when presumptive antimicrobial therapy is commenced. In other words, the decision to treat melioidosis and the choice of antimicrobial agent may have to be taken before any laboratory results are available to guide clinical decision-making. Lengthy courses of treatment are required, including a prolonged eradication phase.

There have been several attempts to formulate clinical guidelines for clinicians faced with patients who might have melioidosis [1,2]. One in particular reviewed the then available literature for evidence of efficacy, and addressed the more difficult dilemmas in clinical therapeutics including duration of treatment, the value of combination therapy, the need for prolonged eradication therapy and the role of other treatment modalities [1]. The key recommendations were use of the cephalosporin Ceftazidime or a carbapenem antibiotic for initial treatment of acute infection over 2-4 weeks and a combination of co-trimoxazole and doxycycline for eradication over a 12-20 week period. Broadly similar recommendations set in the context of diagnostic and clinical review of Australian experience were adapted for use in the emerging disease setting of South and Central America [2]. More recently, those recommendations were updated by an Australian group noted for clinical trials on melioidosis therapy [3]. The key features of these guidelines and other recommendations reviewed here are summarized in Table 1.

The antibiotics used to treat melioidosis fall into two distinct categories: (a) those suitable for the treatment of the acute septicaemia phase of disease (phase 1), and (b) those used in the subsequent eradication phase of therapy, previously known as "maintenance" therapy (phase 2).

The baseline melioidosis mortality rate of 35% reported in Thailand and 19% in Australia has led some centres to consider the use of complementary, non-antibiotic interventions in severe septicaemic infections [1,3,11]. Granulocyte colony stimulating factor (G-CSF) has been tried as an early intervention in severe acute melioidosis in an uncontrolled retrospective study performed in Northern Australia [20]. Disappointing results were demonstrated in a recent randomised controlled trial of G-CSF for the treatment of septicaemic melioidosis in Thailand, although longer survivals were observed in individual cases [21]. Other potential adjunctive agents for which there is less clinical efficacy data include human recombinant activated protein C [22], which has been used in other forms of Gram negative septicaemia, in order to modify the host response to overwhelming infection. There is growing interest in a potential role for statins in the management of acute melioidosis, following the incidental observation that patients with systemic bacterial infection fare better if they receive statin treatment [23].  In vitro studies into statin activity on B. pseudomallei are at an early stage. There is as yet no published clinical experience specific to melioidosis. A recent critical review of clinical trials of statin use in sepsis found a modest and inconsistent effect on clinical outcome [24], and highlights how difficult it will be to assess the impact of this family of non-antibiotic agents on the early stages of a sporadic infection.  It is likely that statins will find a place in the treatment of melioidosis only after a more extensive theoretic and in vitro case has been built for their use and assessed in the setting of randomized clinical trials.

Treatment of melioidosis is a challenge even where there are adequate resources to support patients with multiple organ failure and extensive clinical experience. In resource-poor settings, the cost of optimal Phase 1 and 2 therapy imposes severe restraints and is a likely contributor to unsuccessful clinical outcomes. The duration of the phase 1 of antimicrobial therapy is longer than for most other infections. Pressure on hospital bed occupancy has led to evaluation of a hospital-in-the-home approach which has been successfully used in northern Australia [9]. However, the longer phase 2 of treatment has its own specific challenges, such as compliance and monitoring. Clinical progress assessment and laboratory tests suitable for the first phase of treatment are not as helpful by this stage. C-reactive protein has been used in this role for many years [25], but is unreliable as a sole indicator of imminent relapse [26]. If the earlier phase of initial therapy has been successful, disease will be subclinical; its persistence entirely at a tissue and cellular level. This partly explains the preference some authorities have for longer courses of Phase 2 eradication therapy. The approach taken by the Pahang Melioidosis Registry [17] will in future permit an evaluation of adherence to the 12-20 week phase two treatment and its long term impact on late relapse.

A specific issue raised by the emphasis on biosecurity is a need for post-exposure prophylaxis [27]. Though this has not been required for treatment of first responders to a bioterrorism incident involving B. pseudomallei, there are occasional accidental exposures in the clinical laboratory [28], and during field investigations of outbreaks when high bacterial concentrations may be encountered [29]. Consensus guidelines have been developed and rely on a combination of prompt risk assessment after the event, susceptibility testing of bacterial isolates, post-exposure prophylaxis with one or more of the preferred agents, and careful clinical monitoring [30]. Laboratory experiments with an animal model indicate that there is a 48hr window for administration of post-exposure prophylaxis [31].  However, this has probably done little more than slow disease progression in laboratory animals. There is therefore no reliable evidence to support the use or choice of Phase 0 prophylaxis regimens.

The search for improvements in melioidosis treatment continues with the release of new antimicrobial agents. The new carbapenem Doripenem and the cephalosporin Ceftobiprole have both been shown to have bactericidal activity against B. pseudomallei in laboratory studies [32,33], but have yet to evaluated in clinical trials. Other promising new agents that may have a role in the treatment of melioidosis in future include Ceftalorine and Iclaprim, but assessment of their efficacy is at an earlier stage.

Melioidosis is a difficult infection to manage, not least because of its capacity to cause a rapidly fatal outcome despite the use of appropriate antibiotics. There has been modest progress in the last two decades, though mainly in the details of treatment delivery. At present, a universally effective and inexpensive treatment remains outside our grasp.